the synthesis of condensed tannins declan moran
TRANSCRIPT
THE SYNTHESIS OF CONDENSED TANNINS
DECLAN MORAN BSc.
This thesis is presented for the award of
Masters degree in Chemistry.
Supervisor: Dr. Paraic James
School of Chemical Sciences
Dublin City University
Dublin 9.
Date of Submission: August 1994
The thesis submitted is based upon the
Candidates own work.
by
DECLAN MORAN
I hereby certify that this material, which I now
submit for assessment on the programme of study
leading to the award of Masters degree is entirely
my own work and has not beeii taken from the work
of others save and to the extent that such work has
been cited and acknowledges within the text of my
work.
Signed: Qua.
C a n d i d a t e
TABLE OF CONTENTS:
(i) Abs trac t
C h ap ter 1: L i t e r a t u re R ev iew
T h e Synthes is o f C ondensed Tannins.
C hap te r 2 : Exper im en ta l S e c tion
The Synthes is and Pur if icat ion o f
Procyan id in B3.
A p p e n d ic e s :
(A) H P L C C h ro m a to g ra m s
(B) 400 m H z N M R Spect ra
Page .
1
64
124
125
ABSTRACT:
The Synthesis of Condensed Tannins:- Declan Moran.
A li te ra tu re r ev iew w i th 110 re fe rences is p re sen te d which examines the hi s to ry o f the chem is t ry o f condensed tann ins and the deve lopm en t o f this chemis try in paral lel w i th the deve lopm en t o f m ode rn analyt ical techniques . The role w hich these c o m p o u n d s play in the wine and brewing indus tr ies is discussed . An exam ina t ion o f the di f feren t synthe t ic rou tes applied to the iso la t ion o f c o n d e n s e d tannins, m os t no ta b le p rocyanid in B3, is presen ted . Also inves t iga t ions into the m odes o f ac t ion for the fo rm at ion o f tannins in wine and beers is d iscussed with the implica t ions behind the ir presence explained.
An experimenta l sec t ion tha t exam ines tw o dif ferent synthet ic ro u te s for the fo rmat ion o f procyan id in B3 f rom the s ta r t ing mater ia ls (+)- tax ifol in and (+)- ca techin is p resen ted and the advan tages and d isadvan tages outl ined. N ew analy t ica l m e thods for the m on i to r ing o f these r eac t ions are p resen ted along w i th the N M R spec t ra o f r e co v e re d p ro d u c t s which are explained and c o m p a re d to da ta genera ted f rom l i te ra tu re sources.
A m e thod fo r the ace ty la t ion o f the r eac t ion p roduc ts , thus stabilising them is p resen ted a long with the H PL C and N M R detai ls explain ing the resu lt s obta ined . Final ly tw o sepa ra te a t t em p ts at synthes ising procyan id in B3 via enzymatic ox ida t ion are p resen ted and discussed , and the two m e th o d s are eva lua ted .
ACKNOWLEDGEMENTS:
My sincere thanks are due to my supervisor Dr. Paraic James
for his continuous guidance and encouragement and also to my
employers Bristol-Myers-Squibb Swords Laboratories for their
generous financial support.
I would also like to express my thanks to Guinness Group
Research for their invaluable assistance and for all chemicals supplied.
Finally I would like to express my gratitude to Ann Moran and to
Marie Nally.
THE SYNTHESIS OF CONDENSED TANNINS
INTRODUCTION;
The study o f condensed tannins is among the oldest forms o f chemistry
known. The origin o f the word tannin is derived from the practice for which
these compounds were used in the leather industry. This industry has made use
o f tannins for centuries. Tanning is a process whereby dried animal skins are
formed into leather by its contact with a profusion o f tannins. In the Northern
lati tudes the most common tannin used was oak bark. The reason for this was
the large acreage o f oaks that were a common feature o f all northern European
countr ies in the past. This process had been carried out for generations but
the science behind the expansion o f leather chemistry was born out o f the
Industrial Revolution. For the first t ime leather was being produced for more
than personal use and the age o f mass production had arrived
During the evolution o f this chemistry there have been many pioneers who have
gained a knowledge o f the complex chemical processes involved in tanning.
The attainment o f this knowledge came about very much on a trial and error
basis. Consequently, there have been many definitions which have tried to
encapsulate the true essence o f what defines a tannin. A general description
used defines a tannin as everything extracted from a plant that gives a blue
co lour on exposure to ferric chloride. Haslam in his book on Plant Polyphenols
[] ] prefers the definition coined by Bate-Smith and Swain [2], They proposed
that tannins were "water soluble phenolic compounds having molecular weights
between 500 and 3,000 units and that besides giving the usual phenolic
reactions, they have special properties, such as, the ability to precipitate
alkaloids, gelatin and o ther proteins."
1
In overall terms there are two separate areas that can be discussed in relation
to tannin chemistry. The first has the group o f polyesters based upon gallic
acid and its derivaties, known as hydrolysable tannins and the second contains
the group o f proanthocyanidins, known as condensed tannins. It is the latter
with which this work is concerned Before delving into the field o f condensed
tannin chemistry, it is o f some importance to outl ine many o f the uses to which
these compounds have been applied since the decline o f the leather industry in
the Twenthieth Century.
The initial spark o f interest into these compounds stemmed from the discovery
that they were involved in the defence mechanisms o f many plants. The
astringent nature o f these materials caused an unpalatable taste in the mouths
o f would be preda tors thereby insuring the plants survival. The reason behind
this was identified as being due to the complexation o f proanthocyanidins with
macromolecular proteins such as glycoproteins [3], Allied to this discovery
was the realisation that these compounds also demonst ra ted anti-microbial and
anti-viral properties in plants. Because these compounds are found in the
human diet this was an impor tant discovery. The basis o f many oriental herbal
remedies has been shown to be due to the presence o f proanthocyanidins. For
example these compounds have demonstrated a posit ive effect against
strep tococcus mu tans, the primary causative agent of plaque and dental
cavaties, by limiting the adhesion o f the microbe on the smooth dental surface
[4], As a result many commercially available mouthwashes contain
proanthocyanidins.
2
The anti-fungal properties o f these compounds have been known since the turn
o f the Century when Knudson demonstrated that very few fungi could survive
in the presence o f 2%v/v tannin solutions [5], Studies into the anti-viral
capacity o f these compounds have also been undertaken. Cadman reasoned that
polyphenols complexed with the virus thus making it non-infective [6], All o f
these properties have lead to the increase o f research into these compounds
over the past three decades. The Food and Beverage Industry by extension,
have a major interest in the area o f proanthocyanidins. Since these compounds
are found in numerous plant types, they form part o f the human food chain and
are thus important in this Industry. Polyphenols contribute to the taste and
flavour o f foods and their presence or absence must be controlled. Too much
o f these compounds and the food developes an ast ringent taste. Astringency is
defined as "a feeling o f constrict ion, dryness, roughness, along with a sense o f
stiffness in the movement o f the tongue with some loss o f taste [7],
Within the Food Industry there has been a trend away from synthetic colourants
due to possible toxic effects associated with these chemicals. Anthocyanins are
compounds that are known to possess colouring capability and also may be o f
pharmaceutical importance. The stability o f these compounds is based upon
several factors: pH, temperature , partial oxygen pressure, types o f co-product
present, light radiation and glycosidalion, as well as the nature o f the
heterocyclic rings. The development o f commercial processes to isolate and
purify these compounds is o f great importance. Pifferi et a l have outlined an
optimised procedure which utilises polyvinylpyrrolidone (PVP) and alumina [8J.
This group investigated several classes o f anthocyanins and found that their
method was very effective in separating the various compounds. The alumina
was found to impose a buffering action on the compounds due to the presence
o f both acidic and basic sites thus allowing separation.
3
Since anthocyanins are used as colourants , methods for their identification and
character isation are needed if the strict rules applying to food additives are to
be adhered to. In the United States beverages are only allowed to contain
grape skin extracts while non-beverages are only allowed grape colour extract.
With this in mind regulatory bodies routinely analyse commercial products and
adulteration with banned substances can be identified. The standard way to
measure these compounds was based upon their t inctorial strength (a measure
o f colour efficiency), however, this can be misleading as optical density and
colour rank are poorly corre la ted due to the fact that optical density does not
deal with band widths as does optical perception.
Wrolstad el a / have proposed the use o f high performance liquid
chromatography (HPLC), allied to photodiode array detection (PDA) to
characterise anthocyanins [9], They subjected a whole series o f compounds to
analysis and from results obtained, observed that the method proved to be
useful as the retention characterist ics o f compounds yielded data on the sugar
moieties bound and on the nature o f the anthocyanin. The same group further
developed their technique to build a library o f spec tra for standards mainly
characterising acidic polyphenols such as: delphinidin, cyanidin, petinidin,
pelargonidin, peonidin and malvidin [10]
As well as foods, anthocyanidins contribute to the overall f lavour o f beverages,
most notably teas, wines and beers. In wine production, part icularly for whites
the key aim is to limit the extraction o f polyplienolic material from the grape so
that oxidation of these compounds may be controlled. For red wines this trend
is reversed and the extract ion o f polyphenols is encouraged. The presence o f
such compounds and their subsequent mode o f action is not yet fully
unders tood but major str ides into the unders tanding o f these processes have
been made more recently.
4
Beers and lagers contain small concentra tions o f procyanidins. These are all
derived from the malt (or barley) used in the manufacture o f beer. Their
presence in beer is a two edged sword, on the plus side, they contribute to the
flavour o f the beer, while on the minus side they have been implicated in the
astringency o f beer and also in the formation o f non biological hazes that
shorten the shelf life o f the beer. This is a universal problem for brewers and
although not fully understood, it is known to be caused by the association o f
the procyanidins and other polyphenols with proteins and polypeptides, it is not
known however i f specific proteins are involved.
During the cooling o f a beer chill hazes are formed which are not permanent as
these re-dissolve with heating. For the commercial production o f beer, chilling
is necessary and the haze is an unwelcome problem that must be removed.
Currently one way to remove this haze is to immerse nylon strands into the
beer which cause the polymeric material to physically adhere to it thus
removing the precipitates. Over a period o f time permanent hazes appear in
beers which limit its shelf life. The reason for this has been proposed to be due
to slow acid catalysed (pH 4.0) bond breaking and re-forming reactions
characterist ic o f proanthocyanidins [11]. Under the weakly acidic conditions
that prevail in beers, decomposi t ion o f the proanthocyanidin within the
pre-formed protein-polyphenol complex is thought to occur and this generates a
f lavan-3-ol carbocat ion which is a st rong nucleophilic acceptor open to attack
and polymerasit ion. It has been suggested, however, that polymersition is not
the cause o f haze formation but ra ther that the reactive carbocation is captured
by nucleophilic thiol groups o f proteins. Although these adducts are acid
labile, they are more stable than the proanthocyanidins. Therefore, the protein
surface is more hydrophobic in nature ultimately leading to haze formation.
5
REACTION SCHEME#1 : PROPOSED HAZE FORMATION
MECHANISM.
Some o f the o ther areas that make use o f condensed tannins include the Oil
Industry and the manufacture o f natural adhesives. It has been found that
sulphonat ion o f tannins makes them more soluble and reduces viscosity. This
has been applied to tannins used as lubricants in oil well drilling and adhesive
applications. The reason for this is not yet fully unders tood. Yeap Foo et a l
established that the p rocess only occurred to a small degree [12].
6
T rea tm e n t o f tannins with sodium hydrogen sulphi te lead to the iso la t ion o f
sod ium e p ica te ch in - (4 B ) - s u lp h o n a te (Fig 1) and a dimer sod ium ep ica techin
- (4B -8 ) -ep ica tech in - (4 B ) - s u lp h o n a te (Fig 2) in 20% and 6 % yield respectively.
Result s showed tha t c leavage o f the inter f lavanyl bond by the su lphite increased
solubil ity. This reac t ion should be con t ro l led for lea ther tanning as dimers and
tr imers exhibit po o r tann ing capabil i t ies . Su lphona t ion o f adhes ive tannins also
requires an increase in c ross - l ink ing agen ts to c rea te cured resins.
Fig. I
OH S 0 3Na
Sodium cpicalccliin (4B) sulphonatc
D im er (413-8) - (413) - Sulplioiuito
7
Finally it is w o r th no t ing tha t this r ev ie w is in no way a com ple te accoun t o f all
the areas o f research in to this field car r ied ou t o f the pas t th ree decades . Many
re fe rences detai l the ex t ra c t ion , iso la t ion and charac te r i sa t ion o f tann ins f rom
plant sources . To detai l each o f th e se w o u ld be a mjor u n d e r ta k in g well
beyond the scope o f this rev iew, h o w ev e r , the topics d iscussed shall be direct ly
re la ted to the synthes is o f c o n d e n s e d tann ins from natura l p recu rso rs , the ir
ch a rac te r i sa t ion by ch ro m a to g ra p h ic and sp ec t ro s co p ic m eans w i th in some
cases d irec t com par i son to the i r na tu ra l analogues . Final ly, s tudies tha t have
invo lved mode l so lu t ions tha t in som e w ay mimic the p rocesse s tha t occu r in
n a tu re shall also be ou t l ined in som e detail.
8
THE SYNTHESIS AND CHARACTERISATION
OF PROCYAN I DINS
O ver the pas t th ree decades the re have been many g r o u p s involved in the
s tudy o f procyanid ins . The earl ies t c lass if icat ions o f these com pounds
were a t t r ibu ted to w o r k car r ied o u t in the first tw o decades o f this
century. L e uc oan thocya n id in w as f irst d iscovered as far back as 1915
[13], S tudie s by several dif ferent w o r k e r s showed tha t the se classes o f
co m p o u n d s w ere conf ined mainly to w o o d e d plants (i.e. bark plant types)
[14,15] , Chemical s tudies have sh o w n tha t the re are tw o main types o f
co m p o u n d s in this class. The f irs t type are f lavan-3 ,4 -d io l s and the second
f lavan-3-o l d imers and higher o l igomers , ( f ig .3/4):
Fig. 3
O H
"V
Flu van - 3,4 - diol
O i l
OH
Init ial analysis o f the p rocyan id in d imers o f B type (B1 to B4) was carr ied
out on the ir p e rac e ta te der iva t ives, as these com p o u n d s were more stable
than the free hydroxyl form. H as lam e t a l po s tu la ted tha t i f b iosyn thesis
s tud ie s w ere to be u n d e r ta k en than analysis o f the free form wou ld be
necessa ry [16], H ow e ve r , 'H N uc lea r M agne t ic R eso n an ce (NMR)
t e m p e ra t u r e s tudie s on the se free phenol ic com p o u n d s show ed that the
d im ers B I to B4 exhib ited con fo rm a t iona l i somerism making chemica l shift
a s s igna t ion diff icult due to b road r e s o n an ce signals.
For the procyanidin dimers B3 and B4 satisfactory analysis was possible if
a few assumptions were made. Firstly the lower terminal unit's heterocyclic
ring was assumed to adopt a "skew boat" conformation, with the aryi
group at C-2 in a quasi-axial position. Secondly, based upon the splitting
o f the methylene group o f procyanidin B3, as well as the signal for H-4, it
was assumed that rotat ion about the 6'-4 or 8'-4 interflavanyl linkage was
res tricted rather than having heterocyclic ring flipping. However, due to
the conformational isomerism a third factor is involved, namely, that the
upper ha lf o f the molecule should have 2,3-t rans configuration and the C-3
hydroxy group be in a quasi-equator ia l position.
Degradation studies were carried out to confirm the st ructures that had
been proposed using ethanolic-HCl at 60”C. In all cases the dimers yielded
cyanidin and for B2 and B4 epicatechin, while for B1 and B3 (+)-catechin
(Fig-5/6).
Fig. 5 Fig. 6
(-) - Epicatechin
The group suspected that the interflavanyl linkage was 4-8' in
procyanidins. This was confirmed by the synthesis o f diacetyl
octamethyl-procyanidin. This was a simple cleavage reaction that served
to confirm the link. Arising from this work, studies o f the dimers B5 and
B7 and the tr imers C l and C2 were carried out. The study o f these trimers
was o f significance because it suggested that the controll ing factor in the
1 0
formation o f higher polymeric procyanidins was a chemical ra ther than an
enzymic process. (F ig .7/8)
Fig .7
Oil
Fig .8
OH
no
Procyaiiidin C 1
IIO
Procyanidin C2
In order to improve the study o f these compounds a method for the
determination o f absolute configurations was needed. With the aid o f 'FI
NM R and 13C N M R and also mass spectroscopy the absolute configuration
should be possible to be assigned. Roux et a l used the fact that the
condensation o f f lavan-3-ols at the C-4 posit ion was stereoselective when
linked to resorcinol or phloroglucinol groups, holding partial or total
retention o f 2 ,3-t rans isomers or inversion o f 2,3-cis isomers, to develop a
direct method o f configurational analysis.[17], Here multiple Cotton
effects formed from aryl chrotnophores at C-4 dominated the circular
dichroism spectra (c.d.) o f the cis and trans isomers. This allowed
determination o f the absolute configuration o f this reaction centre.
11
Up until the 1980's much o f the isolation and characterisation o f
procyanidins had been carried out on compounds extracted from various
plant sources. However there was no universal method o f synthesis for
these compounds until Roux el a l presented their findings [18]. This
group generated C-4 carbocations from flavan-3,4 diols and reacted these
with strongly neuleophilic f lavan-3-ol biflavanoids as well as resorcinol
and phloroglucinol. The c.d spectra allowed unambigous identification o f
chiral centres and this lead to a general chiroptical rule that has been
confirmed by other workers [19], The reduction o f (+)-taxifolin with
sodium borohydride, when coupled to phloroglucinol and resorcinol gave
3.4-trans (4.8% yield) and 3,4-cis compounds (4% yield) respectively. The
carbocation formed under mildly acidic conditions was found to be stereo
specifically captured by phloroglucinol and resorcinol to form 2,3-cis and
3.4-t rans aryl f lavan-3-ols analogues containing inversion o f configuration.
(Fig.9/10).
O H
(+) - Taxifolin
O H
l
3,4 -Trans Analogue 3,4 - Cis analogue
It was also observed that there was little compet it ion from self
condensation.
12
The same group extended their findings to the study o f the first
synthesised 2,3-c is-3 ,4-trans aryl f lavan-3-ol by photolytic rearrangement.
They proposed a reaction mechanism for this rearrangement [20]:
REACTION SCHEME#2: PHOTOLYTIC REARRANGEMENT
MECHANISM.
iR O O
R
O R
ihv
O R 1 +
O R
O R
R1 = R3 = H ; R2 = R4 = Oi l
It was thought that these mechanisms needed formal hetrocyclic cleavage
o f the hetrocyclic ether bond with simultaneous intramolecular
recyclisation via a zwitterion by neuleophlic attack o f the phloroglucinol's
hydroxyl group. It was also observed that in some cases inversion around
the C-3-C-4 bond occurred , while in others just inversion about the 3-4
bond was observed.
13
The 2-OH o f the phloroglucinol D ring was in close proximity to the
C - 2 - 0 bond and these would be anti-oriented in an inverted C-2-C-3
semi-chair configuration. This lead to SN2 cleavage o f the heterocyclic
ether and resulted in inversion at C-2. The half chair configuration lead to
the anti-configurational spacing o f the 2 ,3-c is -3 ,4-trans isomer. The
cleavage o f the O-C-2 bond was observed to be anchimerically assisted by
the lone pair electrons o f the 3-OH group o f the phloroglucinol ring, which
formed an oxiran susceptible to nucleophilic a t tack by hydroxyl groups
and thereby leaving the configuration at the C-2 group. Thus under
photolysis, rearrangement must occur to form the more stable group.
REACTION SCHEME #3:INVERS10N/RETENTI0N OF
CONFIGURATION DURING PHOTOLYSIS
(1) Inversion gives the 2,3-t rans-3 ,4-t rans isomer
(2) Retention @ C-2,invers ion @ C-4, gives 2,3-Cis-3 ,4-trans isomer
14
The interflavanyl linkages were still not fully characterised by 'H NMR, as
this relied upon deshielding experiments utilising methoxy derivatives, plus
solvent exchange. Roux ef a l proposed a model that could be used to
explain the linkages. The model was split into three distinct rings.
(F ig-11) [21].
Fig. 11 OlVIe
cMO 7 8 o ' , ' OMe| ® 1 © '
i kI5 4 OR OMc
They used the substi tuted biflavanoids with 6' and 8' linkages to determine
the absolute chemical shifts for the A ring pro tons H-6 and H-8. The
methyle ther derivatives were directly reacted with pyridinium
hydrobromide perbromide. This was applied mostly to the 8' substi tuted
compounds as this site was more nucleophilic and less sterically hindered
than the 6'. The method o f formation was found to follow the trend
8'-bromo followed by 6', 8' di-bromo and finally 2', 6', 8' t r i-bromo
derivatives. The direct bromination o f the 6' site was not possible and this
was thought to be due to steric hinderence. It was felt that the 6' bromo
compound might be generated by partial d e n o m in a t io n o f the 6', 8'
dibromo product. The 6' bromo compound was generated in a 74% yield
using n-butyl lithium at -20°C giving the first synthesis o f the 6 substi tuted
catechin species. This served as the key to the synthesis o f other 6
substi tuted species.
15
Roux et a l used their knowledge to apply a direct biomimetic approach to
the synthesis o f [4,6 and 4,8] linked biflavanoids previously isolated from
commercially important barks. [22], They reacted (+)-mollisacacidin with
(+)-catechin under mildly acidic conditions at ambient temperature. Three
biflavanoids were identified each comprised o f (-)-fisetinidol upper units
and (+)-catechin lower units. The [4,8] all trans isomer predominated with
a 28% yield. (Fig. 12). Next was the [4,6] all trans isomer in a 5.8% yield.
(Fig. 13) and [4,8] 2,3- trans-3,4-c is-2 ' , 3 '- trans isomers in a 16.5% yield,
not easily separated. (Fig. 14).
F ig . 12
O R
[4,8] All lians Isomer
OR
y uOR
[4,6] All trans isomer
•ig. 14
R O O
I
O R
O R R = H. ^ O R
" i V l ^
O R
OR
., \ O R
I OR [4,8] 2,3-trans-3,4-Cis-2',3-trans Isomer
16
Other workers were also studying plant extracts to try and establish and
confirm the l inkage pa tte rns in procyanidins. Hemingway et a l decided to
focus on procyanidin trimers, as there were several combinations o f
configuration possible due to the fact that these compounds were made up
o f more than one procyanidin group, each having three assymetric centres
and a terminal f iavan-3-ol unit possessing two chiral centres[23]. As well
as these parameters the trimers also had either [4,6] or [4,8] interflavanyl
linkages. They made use o f column chromatography to fractionate their
extracts and isolate the desired compounds. Once catogor ised into their
consti tuent groups the compounds were separated from each other by
HPLC. Molecular weight determinations were made using gel permeation
chromatography (GPC). Acid catalysed degradation involving toluene and
thiol yielded 4-benzyl thioepicatechin. (Fig. 15) and (+)- catechin,
implying that the tr imers were composed o f procyanidin units o f 2,3-cis
configuration with (+)-catechin as the flavan-3-ol terminal unit.
Studies involving partial degradat ion gave procyanidin dimers and benzyl
sulphide procyanidins. The presence o f procyanidin B1 was indicative of
[4,8'] linked compounds while B7 indicated [4,6'] compounds.
Desulphurisation followed by reaction with Raney nickel served to confirm
this. This group also proposed a system for naming the compounds based
upon oligosacchar ide classification ra ther than using rigid IUPAC
conventions. The IUPAC method used flavan as the basic ring system and
Fig. 15
OH
SCHi - PI'
17
(RS) nomenclature to define absolute configuration. This was found to be
misleading and difficult to apply to tr imers and higher polymers.
Certain groups have concentra ted their studies on the application o f NMR
to procyanidins. The aim o f this work was to identify key chemical shifts
which helped to define absolute configuration. Por te r ei a l studied
procyanidins o f 2,3-cis conformation using ,3C NMR and found certain
interesting features [24], Firstly these compounds were free from
conformational isomerism at 30HC allowing in terpretation o f their spectra.
The group found that qua ternary carbons a ttached to oxygen in the A ring
could not be dist inguished from one another but were found in the region
154-159 ppm while those o f the pyrocatechol B ring were found near 145
ppm They assigned shifts to the C-6 and C-8 ring carbons, the C-8 being
more shielded. The C-3 and C-5 carbons were broad signals with twice the
linewidth o f any other signal making them easily distinguishable. The
group made use o f proton de-coupling to assign the signals for the
pyrocatechol B ring.
TABLE#!: 13C SHIFTS OF STRUCTURAL SIGNIFICANCE.
Sample Compound Unit * C-2 C-3 C-4 C-6 C-8
Procyanidin B 1 T 76.4 72.6 36.6 96.1 95.6
B 81.7 67.8 28.0 97.0 108.1
Procyanidin B2 T 76.6 72.8 36.7 96.5 96.0
B 79.1 66.2 29.1 97.4 107.7
Procyanidin B5 T 76.6 72.1 37.0 96.9 96.0
B 78.9 66.6 28.9 108.6 96.9
Procyanidin B7 T 76.7 72.0 37.0 96.4 95.9
B 82.0 67.9 28.7 108.4 96.7 [
18
The shifts they obtained were found to be in keeping with those predicted
from the monomer units. Meta -meta coupling distinguished C-6 and C-8
resonances o f upper units from the lower units.
Apart from the extraction o f these compounds from various plant sources,
synthesis had been limited to the coupling o f 4-benzyl thio - flavan 3-ols to
(+)- catechin or epicatechin. However, this synthetic work overlooked the
high reactivity o f leucocyanidins predicted from their structure, based on
the effective delocalisation o f the phloroglucinol type A rings for
4-carbenium ions generated under suitable conditions. One reaction
carried out was the reduction o f (+) - taxifolin by sodium borohydride in
ethanol under n itrogen and its subsequent condensat ion with (+)- catechin
in 1:1 molar ratio. Delcour el a l found 2 .6% o f unreacted (+)-taxifolin
remained with 42% o f unreacted catechin and that the products formed
were two biflavanoids, two triflavanoids and a higher oligomer assumed to
be a tetramer. [25], The [4,8] all trans-bi- (+)-catechin was found in
17.5% yield. This has a spin system Amx superimposed on Abxy o f the
catechin moiety which was indicative o f 2 ,3-t rans-3 ,4-t rans orientation.
This was confirmed by analysis o f the c.d spectra. The 6-H(D) o f the lower
unit had a shift (5 6.16 ppm) that was used to define the 8-substi tution of
the phloroglucinol unit o f (+)- catechin . Other parameters defined for the
perace ta te derivatives were shown to be directly comparable to those
parameters obtained from the natural compound procyanidin B3. The [4,6]
all trans biflavanoid was found in approximately 2% yield and study o f its
octamethyl ether derivative shown that the upfield 8-H(D) (ô6.28ppms)
determined the posit ion o f bonding at C-6 and that the shift difference
between 2-H(F) and 3-H(F)(A60.36) was different to the [4,8] compound and
could therefore be used as a method o f differentiation.
19
This was found to be effective even for [4,8]: [4,6] linked trimers. Other
shift data determined the all trans configuration, while c.d spectra
confirmed the interflavanyl linkage. (Fig. 16).
From parameters defined for the dimers, it was possible to assign the
structures o f the trimers. These were an all trans [4,8]: [4,8] isomer and a
[4.8]:[4,6] analogue in the ratio o f 12:1 respectively and their
concentrat ions approximated those o f the biflavanoids. For the [4,8]:[4,8]
product the 'H NMR spectra showed meta coupled doublets from the A
ring (8 6.11 and 8 6.03ppm) and two overlapping singlets 6-H(D)
(8 6.06ppm) and 6-H(CJ>. These are indicative o f o f the successive [4,8]
coupling of both upper units. The residual chemical shift difference (A8
0.23) also was indicative o f this type o f conf iguration (Fig. 8). Finally the
[4.8]: [4,6] tr imer was differentiated by the deshielding o f one o f it's
singlets 8-H(13) ( 86.06- 86.23ppm) and one o f the two aromatic singlets
relative to one another 6 -H(U) (8 6.07ppm). This showed linkages at 6 and
8 posit ions respectively. The 2-H(I) and 3-H(I) shift difference ( A80.33)
was the same as in the biflavanoid. Again coupling constants suggested an
all trans configuration (Fig. 17).
p : . , i /z on
O H
2 0
ou
ou
14,6 : 4,8J All trans Isomer
H O
H O
\O H
Another group studying procyanidins from the conformational aspect
presented results that served to confirm the findings o f Delcour el al.
Yeap Foo el a l found the conformational isomeric effects in procyanidins
interesting. They observed that the barrier to ro ta tion about the
interflavanyl bond was lower in those compounds where the flavan -3-ol
unit was in pseudoaxial rather than pseudoequatoria l coordination. Most
natural compounds isolated had a configuration o f 2R while those
synthesised showed 2S configuration. From analysis o f synthesised
compounds this group confirmed that these compounds adopted a
sofa-type conformation. [26],
The same group have also used their findings to synthesise the first
branched procyanidin tr imer [27], The acid catalysed condensation of
(+)-mollisacacidin with (+)-catechin yielded the first branched catechin
trimer. The reason for its formation is due to the higher activity o f the
phloroglucinol g roup in relation to the resorcinol group. Thus during the
condensation the interact ion o f the two molecules favoured condensation
to the catechin A ring over the resorcinol group. The suggested
mechanism for formation was thought to be via a quinone methide
intermediate as this rou te was thought to be more effective than the acid
catalysis route.
21
Thus by treating the quinone methide genera ted from (4B) -epicatechin
phenyl sulphide and epicatechin - (4B,6)-(+)-ca techin , the first branched
tr imer was synthesised.
The notion o f quinone methide intermediates had already been discussed by
Hemingway el a l [28], While synthesis o f procyanidins by acid catalysed
condensations had been documented it had only been supposed that
synthesis using quinone methide intermediates was possible. Thus under
basic conditions (pH 9.0) Hemingway el a l reacted (4B)-epicatechin
phenyl sulphide in sodium hydrogen carbonate. They found evidence o f
two dimers after jus t 30 minutes. Therefore it was noted that phenolic
anions reacted faster with quinone methides than the acid catalysed
carbocations.
As can be seen most o f the synthesis work has concentrated on oligomers
with either cis or trans configuration. However, in nature a few examples
o f compounds possessing mixed s tereochemist ry have been repor ted
Delcour el a l have already explained the reasons behind the predominance
o f trans configured compounds [29], They subsequently went on to
condense (+)- leucocyanidin and (-)-epicatechin. Several compounds
resulted from this reaction. First were the posit ional isomers [4,8] and
[4,6] 2 ,3-trans 3,4-trans 2 ' ,3 '-cis-(+)-catechin-(- )-epicatechin
(procyanidins B4, B8). Next a novel [4,8] 2,3-trans 3,4-cis-2 ',3'
-c is -diastereoisomer was identified, along with a novel tr imer [4,8:4,8]
2,3-trans 3,4-trans: 2,3-trans 3,4-trans:2,3-c is-bi -(+)-catechin-
(-)-epicatechin. (Fig. 18).
22
[4,8 : 4,8) 2,3 - trans - 3 ,4 - trans : 2,3 - trans - 3,4
- trans - bi -(+) - calechin - 2,3 - cis - epicalechin
Finally the C-4 epimer o f this tr imer and an all [4,8] all trans tri
(+)-catechin-(-)-epicatechin te tramer were identified The group
concluded that the formation o f compounds o f mixed chemistry was similar
to those formed having (+)-catechin as the nucleophile, the biggest
difference being the formation o f 2 ,3-t rans 3,4-cis upper units. Exactly
why this occurred was not fully understood.
Kolodziej et a l provided the first concre te evidence o f this type o f natural
com pound .[30] They examined ethyl ace ta te extracts o f tormentil
rhizomes. As expected they found procyanidin B3 and its [4,6] analogue
B6 but they also found the previously undiscovered (4B-8) cis analogue
(Fig. 19). Using 'H NMR spectroscopy these compounds could be
distinguished due to coupling constants (J34) and the chemical shifts for
two protons in the lower flavanyl units (5 4.94, 5.0 4,4.38 ppm) for the B3
and B6 cis analogues respectively.
T h e na tura l abundance o f the se c o m p o u n d s was found to be similar to that
o f the synthes ised c o m p o u n d s namely B3: B6:eis = 8:1:1 .
As well as the inves t iga t ions into the synthesis o f the se proeyan id in
co m p o u n d s , major im provem en ts into their analysis have occured.
K archesy e t a / used fast a tom bom bardm en t mass s p e c t ro s c o p y (F A B -M S )
to help s equence p rocyan id ins d im ers and tr imers [31]. This technique
p roved very useful fo r the ability lo be able to d ist inguish be tween linear
and branched (rimers. The types o f c o m p o u n d s that w e re studied included
proeyanid in B d im ers ( B I / B 7 ) , the tr imer p roeyanid in C l , and the
a l te rna t ive d im er proeyan id in A1
n o o n
|4 ,8 | 2,3 trails 3,4 cis hi (+) catochin
(fig 20)
Oil
24
The results o f this work showed FAB-MS to be an excellent technique for
sequencing procyanidins. For these studies all the compounds had
previously been purified chromatographically, but it was felt that the
technique could be applied to sampleswithout prior purification.
25
PROCYANIDINS IN THE BREWING INDUSTRY
As we have already seen the industrial application o f procyanidin chemistry
has been the driving force behind much o f the research carried out over the
past three decades. No where has this been more evident than in the
Brewing Industry. The processes behind haze formation in beer production
has already been discussed and the role o f procyanidins clearly established.
Barley contains all o f the procyanidins found in brewing. Out trup et a l
were among the first workers to characterise the individual procyanidins
present in barley [32], They identified three types o f procyanidin and one
prodelphinidin. Reverse phase HPLC was employed to isolate and purify
these compounds and their characterisation was carried out by 270MHz 'H
NMR o f their acetylated derivatives.
The 'H NMR spectra were made up o f a large number o f overlapping lines
which were assigned to partly overlapping small spin systems. Since
pro tons from the individual rings were found not to interfere, it was
possible to interpret one dimensional spectra under standard conditions. As
with the synthesised compounds some chemical characterist ics were
defined. Firstly all o f the products yielded (+)-catechin after mild acid
catalysis. Secondly, after st ronger acid hydrolysis the order o f elution for
these compounds was established:
D=Dimer
Dl=Delphin id in
D2=Cyanidin
T=Trimer
Tl=Delphin id in
T2=Delphinidin +Cyanidin
T3=Cyanidin + Delphinidin
T4=Cyanidin
26
'H N M R analysis showed broad signals interferring with the sharp aromatic
regions and this was due to the presence o f phenolic OH groups in the free
forms o f the compounds. Interference between the pyran ring and the
aliphatic signals was also observed and finally as was found by other
workers conformational isomerism was observed. The compounds in their
free form were observed to be sensitive to oxidation and hence were
deemed to be unstable. The bottom half o f the molecule was found to be
(+)-catechin while the top half was either (+)-catechin or (+)-gallocatechin.
I f further work investigating the possible reactions o f these materials was
to be undertaken, then their synthesis would be essential as extraction from
plant materials was both inefficient and extremely tedious. Fonknechten et
a l employed the fairly readily available monomers o f (+)-taxifolin
(dihydroquercetin) and (+)-catechin as the precursors o f these synthetic
reactions [33], They proposed the use o f a method which essentially was
an optimised procedure for a reaction first described by Delcour et a l [25],
They suggested that dimers could be isolated in a 50% yield using this
method, an improvement on the 20% yield suggested by Delcour. The
isolated compounds were analysed and the results were compared to those
for natural compounds and were found to be in close agreement.
So far all o f the synthetic work had mainly focused on the procyanidins
with little at tention being paid to the prodelphinidins. Work was presented
at the Phytochemical Society o f Europe International Symposium (1984)
which looked at the biochemistry o f plant phenolics and at which it was
stated that dihydroflavanol was necessary for the synthesis of
prodelphinidin [34], Based upon these findings Delcour et a l proposed a
method for the direct synthesis o f prodelphinidin using (+)-catechin and
(+)-dihydromyricetin as precursors (Fig.26).
2 7
Fig. 26
li "IMO o
I ' : X °H
OH
OH
H O O
(A) (2R,3R) (+) Dihydroquercetin (B) (2R,3R) (+) Dihydromyrcelin
The reaction o f these compounds yielded pyrogallol rings in the upper units
and analysis by ’H N M R o f the acid derivatives o f these compounds
generated some o f the following data [35]:
TABLE#2: 360MHz lH NMR OF PRODELPHIN1DIN B3
DECACETATE
P r o t o n N o . C h c in S h i f t s
(ppm)S i g n a l T y p e S p l i t t i n g
J ( H z )
5 - H CE) 7 . 1 2 d o u b le t 8 .4 0
2 X A r - H (n) 6 . 9 6 s i n g l e t
2 - H (E) 6 .9 1 d o u b le t 2 . 0 0
6 - H (e) 6 . 7 1 d o u b ,d o u b I J I O . 4 0
6 . 6 5 s i n g l e t
8 - H (A)6 - H (A) 6 . 5 0 tr ip le t 2 . 6 0
3 - H (0) 5 .6 1 t r ip le t E J 1 9 . 5 0
2 - H (, , 3 - H (F) 5 . 0 8 - 4 . 9 8 m u l t ip l e t
2 - H (C) 4 . 7 8 d o u b le t 10 .00
4 - H (C) 4 . 5 0 d o u b le t 9 .5 0
4 - H e q (l,, 2 . 9 2 - 2 . 8 4 m u l t ip l e t
4 -H ( p, 2 . 7 1 - 2 . 6 1 m u l t ip l e t
O A c 2 .3 5 , 2 . 2 9 - 2 . 2 4 ( X 6 ) 1 1 X S
O A c 1 . 9 8 , 1 . 9 5 , 1 .7 0 ( X 2 ) 1 1 X S
28
The above da ta show ed tha t the ace ta te was free o f conformationa l
isomerism and the coup l ing cons tan t s speci f ied all t r ans chemis try , which
w ere very close to the values for procyan id in B3. The chemical shift o f the
residual D ring p ro ton (6 6 .65ppm s) served to conf i rm (4B -8)
-prodelph in id in B3 l inkage. (Fig. 27).
The same g roup also identif ied tr imers with both (4B -8 and 4 B -6 ) linked
un it s which they identif ied as procyanid in C2, prode lphin id in C2 and a
tr imer com prised o f (4 ,8 :4 ,8 ) - ( - f - ) -ga l loca tech in- (+) -ca techm -(+)-ca tech in ,
as well as (4 ,8 :4 ,8 ) - (+ ) -c a lec h in - (+ ) -g a l lo ca tech in - (+ ) -ca te ch in .
l idoil
Prodelphinidin 133
29
OTHER AREAS OF INTEREST IN PROCYANIDIN CHEMISTRY.
Matt ice et a l have applied fluoresence decay spectroscopy to try and
establish interflavanyl l inkages. [36], Since 'H NMR could not easily
distinguish between the two conformational isomers observed in these
compounds it was felt that time resolved fluoresence spectroscopy would
be more capable o f determining each o f the isomers relative population.
Measurements for the monomers were made in both water and dioxane
while dioxane only was used for the dimers (due to poor solubility
characterist ics in water) . Excita tion o f epicatechin (4B-8)-catechin and
eipcatechin (4B-6)-catechin was performed at 272; 280 and 292nm giving
an emission band between 310 and 321nm. These wavelengths were used
as the fluoresence quantum yield was independent o f wavelength or solvent
effects.
The time resolved emission for monomers was in accordance to a mono
exponential function and it was found that the fluoresence lifetime o f the
monomer was solvent sensitive. For the dimers the decay values o f their
acetate derivatives showed best approximation to bi-exponential functions.
For comparison the dimer procyanidin A l was analysed as this was known
not to exhibit ro ta tion about the interflavanyl bond. This compound
exhibited a mono-exponentia l function giving credence to the fact that
bi-exponential functionali ty was due to rotational isomerism. The group
concluded that if this was responsible for heterogenicity in fluoresence
decay, than pre-exponent ial factors should reflect the relative populations
o f the major and minor rotamer. These factors were found to be dependent
upon molar extinction co-efficient, the fluoresence emisson spectrum, the
radiative lifetime and the concentra tion o f the fluorescent species.
30
Ferreira el a l investigated the fact that procyanidins that were extracted, or
reacted, at alkaline pH's showed an increased acidity and lower reactivity
towards aldehydes than i f a neutral solvent system was used. They
supposed that catechinic acid type rearrangement products were present.
None o f these types o f compounds had previously been classified. They
utilisied the methodology that had been developed to characterise
phlobatannins on procyanidin B3, in order to obtain these compounds.
They showed that over a period o f 1 Vi hours procyanidin B3 completely
converted to a mixture o f oligomeric procyanidins (30%) and that following
column chromatography four compounds (70%) were identified [37], The
compounds included (+)-catechin (9%) and three C ring modified
compounds. These compounds were characterised as the 8,9-cis
9 ,10-t rans-te trahydropyrano [2,3h] chromene (Fig.28) and 2,3-cis
3,4-trans-4 aryl-2-flavanyl benzopyrans (Fig.29).
31
F'8 28 , 8,9-cis-9,10 Transi tetrahydropyrano-[2,3h] chromene
O R \ / o u 2 OR Minor product
OR A
"s, OR2X iIII =R2 =R3 = H or Ac
F i g . 29
OR / y ^ ,
IL I JOR
OR
OR
3OR
X- ° ^ O R
I , OF?
RI =R2 =R3 = H or Ac
2,3-cis 3,4- Trans-4-aryl-2-ilavanyl benzopyranes
Nuclear O v e rh au s e r Effect d if ference s p e c t ro s c o p y (n .O .e diff) was used to
confirm the c i s - t rans conf igu ra t ions o f the C r ings with assoc ia t ions o f
8 -H (c) and 2 -H (IJ) and 6 - H (n) for Fig. 24 and 2 -H (C) with 2 - I I (1}J and 6 -H (n) for
Fig. 25. The spec tra l da ta for the C r ing isomeric p ro d u c t s was
summarised as fol lows:
32
TABLE#3: C RING ISOMER SPECTRAL DATA.
Proton No. Cliein Shifts
(ppm)
Signal Type Splitting
J ( H z )
2-H(B) 6.82 d o u b le t
6.74 d o u b le t -----
5-H(fl) 6.72 d o u b le t
6-H(B) 6.67 d o u b ,d o u b
6-H(nj 6.65 d o u b ,d o u b
2-H(E) 6.64 d o u b le t
6.25 s i n g l e t
6-H(A) 3-H(A) 6.03 s i n g l e t
8-H(C)2-H(C) 5.58 b r o a d s i n g l e t
3-H(„ 5.33 m u l t ip l e t
9-H(C)3-H(C) 5.20 d o u b ,d o u b
2-HCF) 4.83 d o u b le t
10-H(C) 4-H(C) 4.28 d o u b le t
4-Hax 2.83 d o u b .d o u b
4-H(F) 2.67 m u l t ip l e t
O M c 3.83; 3.82; 3.78;
3.77; 3.76; 3.75;
3.60
A l l
s i n g l e t s
O A c 1.97; 1.91 2 X S
As had been previously observed it was felt that procyanidin B3 would
react via its quinone methide intermediate under basic conditions. The
mechanism for this reaction involves the migration o f the (+)-catechin
moiety, helped by the elect ron releasing phloroglucinol unit at C-4 to
reface and C-2. This is followed by the subsequent pyran recyclisation
through 7 -OH(D) with reface o f the quinone methide.
33
UNDER ALKALINE CONDITIONS.
REACTION SCHEME#5: REACTION OF PROCYANIDIN B3
This reac t ion lead to the invers ion o f abso lu te conf igu ra t ion at C-9/C-3 in
procyanidin . T h e 4 -ayrl -2-f lavanyl benzopyranes were formed due to the
enhanced m igra to ry capacity o f the phlo rog luc ino l unit from C-4 to reface
at C-2. Recyc li sa t ion at 2 - O H {A) and the reface o f the qu inone methide
fo rms the p roduct:
34
REACTION SCHEME#6: FORMATION OF 4-ARYL-2-FLAVANYL
BENZOPYRANES.
T'oh i-HO. !
1on
--- ►
on
on
no__
on/== \
noo
onon
on
on
V.
/O il
’
o
T h e g ro u p o f Ariga e l a / inves t iga ted procyanid ins li .nd B3 from a new
s tand poin t 138], They w an ted to s tudy the an t iox idan t p roper t ie s that
the se co m p o u n d s may possess . It was felt tha t these c o m p o u n d s as well as
inc reas ing the she lf life o f foods might also prevent lipid pe rox ida t ion in
humans . Very little w ork had been done to inves t iga te the se com p o u n d s as
food addi t ives. This w as part ia l ly due to the diff icult ies involved in their
35
isolation and purification. The first step in their analysis was to prepare
authentic samples. The compounds were extracted from legume seeds and
were characterised by ultra violet (UV), infra red (IR) and mass spec (MS).
The compounds that were prepared showed the biggest differences in their
]H N M R spectra. Tables outlining these differences were presented as
follows:
TABLE#4: lH NMR DATA FOR PROCYAN 1DIN Bt.
P r o t o n N o . C h c in S h i f t s
( p p m )
S i g n a l T y p e S p l i t t i n g
J ( I I z )
6 X H (Ar) 7 . 1 0 - 6 . 5 0 m u l t ip l e t
8 -H 6 . 0 2 d o u b le t
2 -H (E) 6 .9 1 d o u b le t 2 . 00
6 -H 6 '-H 5 . 9 5 d o u b le t
2 -H 5 . 0 9 s i n g l e t
2 ' -H 4 . 7 9 d o u b le t 8 . 00
4 - H (Ar) 4 . 6 9 d o u b le t 2 . 0 0
3 '-H 4 . 0 5 m u l t ip l e t
3 -H 3 . 9 8 m u l t ip l e t
3 ,3 ' - O H 3 . 6 2 - 3 . 5 4 m u l t ip l e t
4 ' -H 3 . 1 0 - 2 . 4 0 d o u b ,d o u b
These chemical shifts and coupling compounds were found to be identical
to those values obtained by other workers [39], The table for procyanidin
B3 was also presented:
36
TABLE#5: 'H NMR DATA FOR PROCYANIDIN B3.
P r o t o n N o . C l i c m S h i f t s
( p p m )
S i g n a l T y p e S p l i t t i n g
J ( I l z )
8 X H (Ar) 8 . 5 0 - 7 . 2 0 m u l t ip l e t
9 X H (A0 7 . 1 0 - 5 . 8 0 m u l t ip l e t -----
2 - H 4 .7 3 d o u b le t 8 . 00
2 '-H 4 . 5 5 d o u b le t 3 . 5 0
3 -H 4 . 5 4 - 4 . 2 6 s i n g l e t
4 - H 4 . 3 2 d o u b le t 9 . 5 0
3 '-H 4 . 0 8 m u l t i p l e t
3 , 3 ' - O H 3 . 8 3 - 3 . 7 2 m u l t i p l e t
4 ' -H d o u b .d o u b
This data was found to agree with the values obtained by other workers
[6].
Both o f these compounds were selected to be tested in model oxidative
studies. This involved testing their ability to bleach 13-carotene in the
presence o f linoleic acid. This bleaching was due to oxidation.
a -T o c o p h e ro l is a known anti oxidant and was used as a control by which
the effectiveness o f the procyanidins was compared. Since these
compounds have been shown to be good hydrogen donors the same group
set about determining how effective they might be as radical inhibitors [40],
This is important as many reports have suggested that lipid peroxidation
may be linked to cancers [41,42],
37
The te chn iques deve loped by Y a m a m o to e t a l [ 42] as well as m easur ing
oxygen u p ta k e w ere found to be g o o d quan t i ta t ive m easu re s fo r the
de te rm ina t ion o f the level o f lipid p e ro x id a t io n , r a th e r than the m o re usua l
subs t ra te level de te rmina t ions . The p rocyan id in B d imers w ere show n to be
ef fect ive rad ical scavenge rs against pe roxyl radicals and tha t each dimer
had the abili ty to t r ap 8 peroxyl rad icals . A genera l m echan ism for
phenolic an t iox idan ts w as p re s e n te d and i ts m o d e o f ac t ion d iscussed with
pa r t i cu la r r e fe rence to procyan id in B1 and B3. S tudies show ed the m ode
o f ac t ion fo r s caveng ing for B3 to be rela t ive ly slow. Procyan id in B3 w as
found no t to b e ef fec t ive against all types o f com pounds .
3 8
THE SYNTHESIS AND CHARACTERISATION OF OTHER
PROANTHOCYANIDINS
As well as the work involved in the characterisation o f procyanidins that
has already been discussed, it is worth mentioning some o f the studies
involving other related proanthocyanidins such as fisetinidins and
robinetinidins. These compounds were first isolated in the 1960's by
Drewes el a l , who characterised them, in their derivative form, once
extracted from black wattle bark [43,44], For many years the absolute
configurations o f these compounds were open to question. Roux el a l
applied the techniques developed in the study o f procyanidins to finally
determine the configurations o f these compounds [45], The condensation
o f (+)-mollasacacidin and (+)-catechin yielded 3 dimers in 28, 16 and 5.5%
yields respectively (Fig. 29-31).
Fig 29
Fig 30
(-)-fisctinidol - (+) - Catechin(-)- Robinetinidol - (+) - Catechin
39
Fig 31
H
OU
'OH
HO J L OH
V
O (4B - 6) - Robinetinidol - (+) Catechin
OH
OH
By comparison o f these synthetic compounds with those isolated naturally
it was concluded that in nature 2,3-trans 3,4-cis diastereoisomers
predominated, while synthetically the all trans isomer was dominant. Roux
et a l by extension found that trimers existed in extracts o f black wattle bark
[46], These compounds were found to be angular [4,6:4,8]
prorobinetinidin trif lavanoids and their absolute configurations were
established (Fig. 32). The same group were also interested in tetrameric
compounds, which they felt should also be present. In order to confirm this
they exclusively reacted the resorcinol flavanyl units. The reaction o f
(+)-mollisacacidin and (-)-fisetinidol is a stereochemically selective
react ion giving a good yield. This react ion yielded the expected all trans
biflavanoid and its cis analogue, a new linear triflavanoid and the first
branched te traflavanoid to be synthesised [47], 'H NMR showed the
tetraflavanoid to be linked to the triflavanoid, having [4,6:4,8:4,6] linkages
(Fig. 33).
40
Fig 32
Fig 33
.Oil
no o
I
on
o n,on
no
noÖ -
J Uon
on
'OH
ö -on
o nOil on
14,6 : 4,8] A ll Trans bi - (-) Robincliuidol - (+) - Gallocatechin
HO
C v /L 1 1
,OH
. ÙOH
OH
HO ~ f A
' Ï
pHSNOH rA yO II. «* Q O H
Oil
"OH [4,6 : 4,8 : 4,6] - Tetramer
r \W 0,1
0,1 OH
H e /
H O OH
ul À
I!
OHOH
41
The iso la t ion and ident if ica t ion o f na tu ra l ly occur r ing r ing isomer ised
prof ise t inid ins p ro m p te d re sea rch into the ir synthes is [48]. The reason for
this in teres t was due to the usefu lness as co ld set adhes ives tha t the se
com p o u n d s possess. R o u x e l a l rea l i sed the se c o m p o u n d s had the ability to
behave as adhes ives th ro u g h the l ibera t ion o f the reac t ive nucleophi l ic
resorc ino l u n i t s . [49], This was poss ib le i f ep imer isa t ion at C-2 o f
(+) -ca tech in via the in te rm ed ia te fus ion o f the he te rocyc l ic r ing to o k place
[50], U nder bas ic cond i t ions R oux e l a l found 5 s te reospec if ic isomers
f rom the ( - ) - f i se t in ido l - (+ ) -ca tech in p recu rso r . The first was the expec ted
phloba tannin , a p r o d u c t o f the C r ing i som er isa t ion o f the paren t co m p o u n d
(Fig. 34). The second was its C-2 (F r ing) ep imer , w he re (+) -ca tech in has
been con v e r ted into (+) -ep ica tech in . The th ird and fou r th w ere pos i t ional
i somers o f the f irs t and second, w he re (- ) -f iset inido l w as pos i t ioned at the
C-6 pos i t ion o f the (+ ) -ca tech in reac t ing th ro u g h its 5 -O H group. The fifth
w as a s t ruc tu ra l i som er showing an a l te rna t ive m o d e o f cycl isat ion with the
7 -O H g roup (Fig 35).
Fig 34:
1 , 0 ^ OHPhlobatannin
42
Fig 35HO
n o o
u ^ f -
HO\
Y 'OH
o
• ' ' ^ i v O H
Oil
^ 0 1 1
7 - OH Cyclised Isomer
Roux and oil ier c o - w o r k e r s also inves t iga ted oxida tive coupl ing involving
f lavan-3-o l u n i t s . [51]. Coupl ing involv ing l lavones and f lavanones was
well do cu m en ted [52] but, coup l ing involving f lanan-3-o ls was not
common. All know n reac t ions o f this type have involved 2 '-8 l inkages o f
(+ ) -ca tech in via ihe re spec t ive 13 and A ring yie lding biphenyl
d ehyd rod ica tech ins [53]. The d ehydrod ica tech in s were p repared using
enzymatic ox ida t ion or ex t rac ted from black teas af te r fe rmenta t ive
perox ida t ion [54], Weinges e ( a ! s uppose d tha l (+ ) -m esqu i to l (Fig. 36)
would m ore readily u ndergo ox ida t ive coupl ing Ilian ca tech in due to ils
more ac t ive A ring system.
Fig 36:
( i ) - Mesquitol
43
Synthes is o f this co m p o u n d with (+ ) -ca tech in lead to the i so la t ion o f a
[5,6] and [5,8 ] d im er and a [5 ,6 :5 ,8 ]b i s - (+ ) -m esq u i to l - (+ ) -ca tech in tr imer
(Fig. 37). This ind ica ted an a l te rna t ive fo rm o f synthesis to those already
m ent ioned us ing ox ida t ive phenol coupling.
Fig 37:
OH
OH
i T\ ^ O H
^ " » F I
o
\
//^ O H
O H
[5,6 : 5,8] - bis - (+) - Mesiquitol - (+) - Catechin
O th e r w o rk e r s w e re also in te res ted in the base catalysis reac t ions o f
p roan thocyan ins . L aks e t a / [5 5 ] set a bou t veri fy ing the f indings o f Sears e t
a / [5 6 ] who cla imed tha t base r eac t ions lead to an int ra rea r r ange m en t o f
(+ ) -ca tech in yielding ca techin ic acid and isoca tech in ic acid. This acid was
shown to be an enol ic fo rm o f (+ ) -ca tech in - (+ ) -ph lo rog luc ino l (F ig .38)
44
Fig 3 8 : pHI K X ^ ^ O H
X
u o ^ o
6 - (3,4 - dihydroxyphenyl) - 7 - hydroxy - bicyclo[3.3.1] - nonane - 2,4,9 - trione
Since this o c c u r r e d for (+ ) -ca tech in it was p o s tu la ted tha t it may occur in
h igher o l igomer ic c om pounds . P o lymer ic p roan thocyan ins w e re reac ted
with ph lorog luc ino l at pH 12.0 at 23°C and 50°C. It was found tha t bo th
the inter f lavanyl bond and the pyrane e the r u n d e rw e n t rapid c leavage and
tha t the l ibe rated lo w er ca tech in unit w as fu r th e r cleaved at the pyran r ing
Fig 39 :H O v / ^ r O H
Catechinic acid
to form a reac t ive qu inone meth ide tha t in t ram olecu la r ly rea r ranged to form
ca techin ic acid (F ig .39). The adduc t fo rm ed f rom this r eac t ion was not
isola ted .
F e r re i ra e t a l w e re in te res ted in the base catalysed r eac t ion o f
ph loba tann ins due to the i r m ore soluble charac te r i s t ics [57], The need to
d issolve the se c o m p o u n d s has al ready been outl ined [58,59] and is very
im por tan t for the i r prac t ica l handl ing in industr ial s ituat ions.
45
The base ca ta lysed r eac t ion o f (- )- f iset in idol [4G-8]- (+) -ca tech in gave fou r
p roduc ts . These w e re the 8 ,9-c is 9 , 10- trans t e t r a h y d ro -2 H ,8 H -p y ra n o
[2 ,3h] chromene (F ig .40) and its all t r ans ana logue as well as the isomer ic
pairs o f these c om pounds . This g roup w en t on to detai l the absolu te
conf igura t ion for the se c o m p o u n d s and also p re sen te d in form ation on the
[413-6] analogues . F e r re i ra e t a l also de te rm ined the s t ruc tu re s o f base
ca ta lysed c o m p o u n d s o f ( - ) - f i se t in ido l - (+ ) -ca tech in tha t had 2 ,3 - t rans
3 ,4-c is f lavan-3-o l con s t i tu te n t units [60], Final ly this g roup also published
s t ruc tu re s for the f irs t prof iset in id ins and p rogu ibou r t in id in s based on C-8
subs t i tu ted (- ) - f ise t in idol uni ts and the ir C8 ra la ted isomeric com pounds
[61]-
Fig 40 :
8,9 - cis - 9,10 - trans tetrahydro - 2H,8H pyrano [2,3h] chromene
H as lam e l a l [63] have studied the ro le o f p roan thocyan id in s in re la t ion to
the p igm enta t ion o f f low er ing plants. These com p o u n d s have been shown
to be stable under ac id ic cond i t ions since they can exist as f lavylium ions.
H o w e v e r , when the pH is inc re ased d e p ro to n a t io n is apparen t and the
a nhydrobase and its an ion result . This is r ep re s en ted by a c o lou r change
f ro m red to blue .
46
The p roan thocyan id in on its own cannot effec t this co lou r change and this
leads to the th eo ry o f co -p igm en ta t ion , w he re the p resence o f po lypep tides
and po lysacchar ides is implicated in this reac t ion [63], F u r the rm ore , this
g ro u p ascer ta ined som e o the r key p a ram e te r s in plant co loura t ion [64],
The variab les involved were p ro p o s e d to include: A n th ocyan in -co -p igm en t
type , pH, t e m pera tu re , conce n t ra t ion and the type o f metal salt present .
M any po ten tia l co -p igm en ts have been ident if ied from flavanoyl and
hydroxy cinnamyl e s te r s [65 ,66 ,67] , but no de ta i led quan t i ta t ive analysis
was presented . To this end H as lam e l a l used vege tab le tannins , caffeine,
theophyline , adenos ine t r i -p h o sp h a te (A T P ) , deoxyr ibose nucleic acid
(D N A ) and r ibose nuc le ic acid (R N A ) and these were associa ted with
malvin ch loride and cyanin chloride. The ba th o ch ro m ic shif ts were
es tab l ished for these com pounds . The p h e n o m e n u m o f an thocyan in -co
-p igm en ta t ion may be bes t exp la ined in te rm s o f s imilar in terac t ions tha t
are fo rced in an a q u e o u s media by h yd rophob ic effects . There fore , in a
na tu ra l con tex t phenol ic es te rs may be seen as e lec t ron r ich systems
capable o f assoc ia t ion with the e lec tron defic ient f lavylium cation.
Finally, many w o rk e r s have p ro p o s e d tha t the type o f binding involved
b e tw ee n p roan th o c y an in s and p ro te ins is in fac t h yd rogen bonding
[68 ,69 ,70 ,71 ,72 ] , A r tz e l a l used the in te rac t ion be tw ee n synthet ic
p roan thocyan id in d im ers and tr i iners and bovine serum albumin (BSA) to
conf irm this fact [73], Thei r s tudies , h ow ever , also d em o n s t r a ted tha t
hydrophob ic assoc ia t ions were also respons ib le for some binding in many
cases.
47
SOLUTIONS
POLYPHENOL REACTIONS IN MODEL AND NATURAL
In na tu re the types o f r eac t ions tha t have been d icussed tend to take par t in
a reas o f plants tha t are low in enzyme act ivity. There is, how ever , a
second mechanism in na tu re by which p roan thocyan ins reac t . The act ion o f
ox idases gen e ra te po lymer ic mate r ia ls tha t differ f rom each o the r by vir tue
o f the pos i t ion and n u m b e r o f hydroxyl g ro u p s on the molecu le , upon which
the oxidase may act. The resu l tan t p ro d u c t s are assumed to be o f tw o basic
s t ruc tu ra l types. The f irs t are the g roup A procyan id ins which have the
empirical formula C 30H 24O ]2 and the second, the g roup B procyan id ins with
the formula C 30H 2f)O , 2. The enzymat ic d eh y d ro g en a t io n o f ca tech in is an
im por tan t reac t ion in na tu re by v i r tue o f its c on t r ibu t ion to such p rocesses
as co lo u r fo rm at ion and f lavours .
W einges e t a l used ox ida t ive enzymes (pe rox idase , laccase, ty ros inase) to
s tudy the p ro d u c t s fo rmed us ing ca techin as the p recu r s o r [74], They
found tha t in aqueous sys tems the varia t ion and c o n ce n t ra t i o n and dura t ion
o f exposu re o f ca tech in to the enzyme d irec t ly de te rm ined the com posi t ion
o f the p ro d u c t fo rmed . F o r example, i f a f te r tw o days the reac t ion was
quenche d the resu l t ing p ro d u c t w as found to be 8 -hydroxy-ca tech in , while
af te r 10 to 14 days a ye l low prec ip i ta te fo rm ed which when iso la ted and
pur if ied was assumed to be d ehyd rod ic a tec h in A (Fig. 41). The same g roup
fully ass igned the conf igu ra t ion o f this co m p o u n d using fu r ther
sp ec t ro sco p ic te chn iques to study the b ro m o h ep tam e th y l e the r der iva t ives
[75], The final p r o o f was p rov ided by Van Soes t w ho conf irmed the
s t ru c tu re by X-ray analysis o f the b rom o der iva t ive [76],
48
Fig 41 :
HO O H/
HO
HO \
OH *o
\2O
" X J ' ■S'HO
(a) Composition # 1
II2
jOH
O il
'> H t v — \ ° "t i n / OH O -
W / /
OH
The effect of enzymes in natural products is dependent upon several
parameters. The enzymic browning o f grapes is a well studied example o f
these parameters [77,78], where a limiting fac tor in browning is said to be
the oxidisable substrate. Phenolic compounds such as the flavans have
been found by many workers to be important substrates in enzymatic
oxidation due to their capacity to be broken down by phenol oxidases
[79,80,81], Since these compounds have been shown to effect wine during
its production much interest into the nature o f these reactions has been
expressed. Oszmianski et a l s tudied the effects o f phenol oxidases on 7
compounds and monitored the reactions by HPLC [82], Results showed
that degradation o f all compounds occurred but there was, however, no
evidence o f new peaks de tec ted by HPLC. This was in keeping with the
findings o f other groups [83], Allied to the disappearance o f the input
materials, was the presence o f considerable browning.
49
The breakdown rates were observed to be different for different substrates.
To some extent this could be explained by competit ive reactions or,
alternatively, a synergistic effect may help cause dedgradation and this has
been reported in studies involving beers [84], This was thought to occur
from coupled chemical oxidation with quinones formed from enzymatic
reactions. I f this is the case then enzyme oxidation may not, in fact, be the
limiting factor in browning, the enzyme being exhausted early on leaving
any further browning to occur by non-enzymatic means.
The total content o f phenols in a wine is a rough measure o f the capacity
that wine has for oxygen uptake , its ability to withstand oxidation and its
capacity to change when exposed to oxygen. Even though the browning
process is thought to be non-enzymic it is certain that enzymes have a part
to play in oxidation. A good review o f this area was presented by
Singleton who pointed out some o f the practical implications for
polyphenols with specific reference to wine [85], All o f the points put
forward in this review were fur ther upheld by the work o f Cheynier el a l
who studied the effects o f grape polyphenol oxidase on several pheolic
compounds in model solutions [86], They concluded that oxidative
polymerisation leading to the formation o f brown pigments depended upon
the nature and relevative concentra tion o f the phenolic compounds present.
The same group also studied the effects o f trans caftaric acid and
2-S-glutathionyl caftaric acid in model solutions. They found that the rates
o f oxygen uptake very much depended upon the ratios and concentrations
o f the substrates present and concluded that this may account for the
different browning potential among various grape varieties [87],
Fur thermore , this group developed a derivatisation technique involving
benzene sulphinilic acid in order to study the quinone formation via
enzymatic reactions [88],
50
Finally, they studied oxidation kinetics in enzymatic reactions and found
that procyanidins were not directly susceptible to enzymatic oxidation but
they are oxidised by the O-quinones genera ted by caffeoyltartaric acid.
They demonst rated that the coupled oxidation o f procyanidin products
regenera ted caffeoyltartaric acid from its quinone [89],
Oszmianski et a! further investigated enzyme oxidised products o f catechin
and chlorogenic acid at various pH's [90], They made use o f HPLC-PDA
detection to study the formation o f dimers and higher oligmers and the
co-polymers formed from a mixture o f these compounds. Based on earlier
work [91,92] the group proposed that one o f the dimers produced was
procyanidin B3 although no comfirmation o f this was offered. The same
group extended their studies to encompass the effect o f oxidation of
phloroetin glucoside on chlorogenic acid and catechin in model solutions
[93], Previous work had shown that phloroetin glucoside reacted very
slowly to oxidation on its own but in a mixture containing epicatechin a
synergestic effect was observed [94], This was confirmed when it was
shown that the rate o f oxidation increased significently after addition o f
catechin or chlorogenic acid to a phloroetin solution.
Having established that browning processes in fruit and beverages is both
enzymic and non-enzymic its practical implications must be considered.
The browning o f fruit tissue during or after harvesting is a major factor
leading to a loss in quality and yield [95], Therefore, for fruit processing
the control o f this problem has always been a difficult problem to overcome
[96], Many chemicals have been proposed as inhibitors to browning, halide
compounds and aromatic carboxylic acids are known to inhibit polyphenol
oxidases [97], while compounds from ascorbic acid and dextrin derivatives
have also been shown to be effective [98],
51
Sulphites have also been shown to be effective although their popularity is
somewhat curtailed due to the possibil ity o f toxic side effects [99,100],
Methods for the use o f stabilised ascorbic acid or cyclodextrins have been
proposed [101], Very few studies into the use o f thiol compounds, such as
cysteine, have been undertaken although their usefullness has been long
established [102], The mode o f action o f cysteine was proposed as being
via the formation o f colourless addit ion complexes with the forming
O-quinones [103], Very few o f these complexes have been structurally
identified [104,105], and as a result Richard et a l developed a rapid method
for the preparation and purification o f cysteine adduct complexes for
different phenols [106], The formation o f such complexes was monitored
by HPLC and their st ructures were identified by JH COSY NMR. For
catechins it was shown that two conjugates appeared, one at the 2' posit ion
o f the B ring and the other at the 5' posit ion o f the B ring.
Finally, a recent reference has cited the notion o f using enzymes to
reg io -pro tec t and de-pro tect catechin, as a means o f being able to select the
site o f a t tack in further reactions [107], Catechin and its derivatives have
already been demonst ra ted to have impor tant medically applications,
epatoprotec tive [108], anti-choiesteremic [109] and anti-neoplastic [110].
They have also been used as artificial sweetners [111] and as a natural base
for cosmetic products. During reactions it was noted that the B ring
remained free from attack, therefore, in order to preferentially protect this
group it was necessary to perform a hydrolysis rather than an esterification.
Catechin however was shown to have poor solulility in aqueous solutions
so a bio-catalysed alcoholysis was developed using 1-butanol in THF and
an immobilised enzyme. From this two partially acetylated compounds
were isolated (Fig. 42).
52
Fig 42 :
HOAc
iSS i^ O A c
OHAcO
HO
HO
(i) 3,3',4',5 - O - triacetyl Catechin (ii) 3,3',4' triacetyl Catechin
CONCLUSION:
From all o f the work that has been examined in this review, a few striking
features dominate. Firstly some o f the earliest references date back as far
as the turn o f the century, indicating the importance with which this
chemistry was viewed from an industrial standpoint. Secondly much o f the
synthetic work was derived from a need to be able to assign the absolute
configurations o f those compounds that were isolated from natural sources.
Thirdly and perhaps most significantly a glance at the references serves to
show the relatively few workers involved in this area. This would appear
to be a good indicator as to the difficulty involved in the isolation and
purification o f such compounds.
Finally as was stated in the in troduct ion to this publication, the body o f
work presented in no way is an accurate review o f the entire area o f
proanthocyanidin chemistry. Rather the work has mainly concentra ted on
the synthesis o f these compounds, leaving the large area o f extractions
from natural sources untouched to a greater extent. With the growing
pressures involved in the use o f more natural food additives, allied to the
industrial applications, it would appear that the analysis o f these
compounds should continue to flourish in the future.
REFERENCE LIST :
[1] H as lam E.; i n P l a n t P o l y p h e n o l s , C am b r id g e Univers i ty Press, 1989.
[2] Ba te -Smith E .C . ,S w a in T. ; in C o m p a r i t i v e B i o c h e m i s t r y ,Eds. Mason
H.S, Flork in A M., vol 3, New Y ork A cadem ic Press, (1962) , 764.
[3] B a te -S m i th E .C . ; P h y t o c h e m i s t r y , 12, (1973 ) , 907.
[4] K ak iuch i N. , H a t t o r i M., N ish izaw a M., Yanag ish i T. , O kuda
T . ,N am ba T. , C h e m . P h a r m . B u l l . , 34 , (1986) , 720.
[5] Knudson L . ; . / . B i o l . C h e m . , 14, (1913) , 159.
[6] Cadm an C .H. ; in P l a n t s i n H e a l t h a n d D i s e a s e , E d .P r idham J .D.,
Oxford and L o n d o n , P e rg a m o n P ress , (1960 ) , 101.
[7] B a te -S m i th E .C . , Swain T.; C h e m . a n d I n d . , (1953) , 377.
[8] Pi fieri P .G., Vaccari A . , . / . F d . T e c h n o / . , 16, (1981) , 283.
[9] Wrols tad R.E., H o n g V .; . / . A g r i c . h 'o o d . C h e m . , 38, (1990) , 698,
[10] W rols tad R.E., H ong V.;. / . A g r i c . h o o d . C h e m . , 38, (1990) , 708.
[11] B ea r t J .E . , Li l ley T .H . , Has lam E.; J , C h e m . S o c . P e r k i n T r a n s . 2 ,
( 1985 ) , 1439.
[12] Y eap F o o L., M c G r a w G.W . , H e m in g w a y R .W .; J . C h e m . S o c .
C h e m . C o m m u n . , (1983 ) , 672.
5 4
[13 I Rosenhe im O., B i o c h e m . 14, ( 1920 ) , 178.
[14] Ba te -Sm i th E C ; . / . E x p . H o t . , 4, (1953 ) , I.
[15] Ba te -S m ith B.C., L erne r N .H . ; B i o c h e m . 58, (1954) , 126.
[16] Has lam H., T h o m p so n R.S . , J acq u es D., Tanne r R . J . N . ; ./. G h e n t .S o c .
P e r k i n T r a n s / , (1972) , 1387.
[ 17J R oux D.G., B o th a J .J ., Fe r re i ra D.; J . C l i e n t . S o c . C l i e n t . C o m m u n
(1978) , 698.
[18] Roux D.G., B o tha J.J. , Y oung D.A ., Fer re ir a D . ; ./. C h e m . S o c .
P e r k i n T r a n s . 1 , (1981) , 1213.
[19] H as lam E. , B a r re t t M .W ., K lyne W., Scopes P.M., P o r t e r L.J . ;
J . C h e m . S o c . P e r k i n T r a n s . / . , ( 1979 ) , 2375.
[20] Roux D.G ., van der W es thu izen J . l l . , F er re i ra D . ; ./. C h e m . S o c .
P e r k i n T r a n s . / , (1981) , 1220.
[21] Roux D .G . , H und t H .K .L . ; . / . C h e m . S o c . P e r k i n T r a n s . 1 , (1981),
1227.
[22] Roux D.G ., B o th a J .J . , Y oung D.A ., Ferrei ra D . ; .) . C h e m . S o c .
P e r k i n T r a n s . / , (1981) , 1235.
[23] H em ingw ay R .W . , Yeap F oo L., P o r t e r L . J . ; C h e m . S o c . P e r k i n
T r a n s . ( 1982 ) , 1209.
55
[24] P o r te r L.J., N ew m an R .H . , Yeap Foo L. , W ong H., H emingway
R.H.; «7. C h e n t . S o c . P e r k i n T r a n s . (1982 ) , 1217.
[25] D elcour J. A., Fer re ira D., Roux D G . ; J . ( 'h e m . S o c . P e r k i n I r a n s .
Z., (1983) , 1711.
[26] Yeap Foo L., P o r te r L J . ; . / . C h e n t . S o c . P e r k i n T r a n s . / . , (1983) ,
1535.
[27] Yeap Foo L., H em ingw ay R .W . ; . / . C h e m . S o c . C h e m . C o m m u n . ,
(1984) , 85.
[28] H emingway R. W., Yeap Foo L.; J . C h e m . S o c . C h e n t . C o m m u n .,
(1983) , 1035.
[29] D e lcou r J A., Serneels E.J . , Fer re ir a D., Roux D.G .; ./. C h e m . S o c .
P e r k i n T r a n s . 1 . , (1985 ) , 669.
[30] Kolodz ie j H. , Seh leep S., Fried r ich H .; J . C h e n t . S o c . C h e m .
C o m m u n . , ( 1 9 8 6 ) , 392
[31] Karchesy J .J . , H em ingw ay R.W., Y eap Foo L., Barofsky E. ,
Baro fsky D.F .; A n a l . C h e m . , 58, (1986 ) ,2567 .
[32] O u t t ru p M., S c h a u m b u rg K.; C a r l s b e r g R e s . C o m m u n . , 46, (1981) ,
43.
[33] F onknech ten G., Moll M., Cagnian t D. , Kirsch G., Muller J .F. ;
./. I n s / . B r e w . , 89 (1983 ) , 424.
56
[34] Outtrup H.; The P h y tochem ica l Soc ie ty o f Europe, In tern a tio n a l
Sym posium , Ghent, (1984).
[35] Delcour J.A., Vercruysse S. A.R.; J. Inst. B rew ., 92, (1986), 244.
[36] Mattice W.L., Bergmann W.U., Barkley M.D., Hemingway R.W.;
J. Am. Chem. Soc., 109, (1987), 6614.
[37] Ferreira D., Steynberg J.P. , Bezuidenhoudt B.C.B., Burger J .F.W.,
Young D.A.; J. Chem. Soc. P erkin Trans. I , (1990), 203.
[38] Ariga T., Koshiyama I., Fukushima D.; Agric. B iol. Chem ., 52,
(1988), 2717.
[39] Nonaka G., Nishioka I., Nagasawa T., Oura H.; Chem. Pharm. B u ll.,
29, (1981), 2862
[40] Ariga T., Hamano M.; A gric. B iol. C hem ., 54, (1990), 2499.
[41] Player T.; in Free R a d ica ls L ip id P eroxida tion an d C ancer , Eds.
MeBrien D C .H., Slater T .F . ,Academic Press London (1982), 173
[42] Osawa T., Ide A., Su J.D., Namiki M.; J. Agric. F ood Chem .,
35, (1987), 808.
[43] Drewes S.E., Roux D.G., Saayman H.M., Eggers S.H., Feeney J.,
./. Chem. Soc ., (1967), 1302.
[44] Drewes S.E., Roux D.G. Eggers S.H., Feeney J.; J. Chem. Soc.,
(1967), 1217.
57
[45] Roux D.G., Bolha J.J., Ferreira D.; J. Chem. Soc. Chem. Com mun.,
(1978), 700.
[46] Roux D.G., Viviers P.M., Botha J.J., Ferreira D., Saayman H.M.;
Chem. Soc. P erkin Trans. 1, (1983), 17.
[47] Roux D.G., Steenkamp J.A., Ferreira D., Hull W.E.; J. Chem. Soc.
P erkin Trans. 1, (1983), 23.
[48] Steenkamp J.A., Steynberg J.P., Brandt E.V. , Ferreira D., Roux
D.G.; 7. Chem. Soc. Chem. C om m un., (1985), 1678.
[49] Roux D.G., Steynberg J.P., Young D.A., Burger J .F.W., Ferreira D.;
J. Chem. Soc. Chem. C om m un., (1986), 1013.
[50] Freudenberg K., P u r rm an n L . ; Chem. B er., 56, (1923), 1185.
[51] Roux D.G., Young E., Young D.A., Ferrei ra D.; ./. Chem. Soc.
P erkin Trans. 1, (1986), 1737.
[52] Locksley H.D.; Forischr. Chem. Org. N a turst., 30, (1973), 207
[53] Weinges K., Bahr W., Ebert W., Goritz K., Marx H .D .,Forischr.
Chem. Org. N a tu rst., 27, (1969), 158.
[54] Vuatez L., Brandenberger H.; J. C hrom atogr., 5, (1961), 17
[55] Hemingway R.W.,Laks P.E., Conner A.H.; J. Chem. Soc. Perkin
Trans. 1, (1987) , 1875.
58
[56] Sears K.D., Casebier R.L., Herger t FI.L.,Stoltz G.H.; J. Org. Chem .,
39, (1974), 3244.
[57] Ferreira D., Steynberg J.P., Young D.A., Brandt E.V., Steenkamp
J A . ; J. Chem. Soc. Chem. C om m un., (1988), 1055.
[58] Pizzi A., in W ood A dhesives: C hem istry an d Technology, Marcel
Dekker, New York, (1983).
[59] Kreibich R.E., Hemingway R.W.; P roceed ings o f IU F R O N -TR I
Sym posium on W ood A dhesives, 17, (1985), 1.
[60] Ferreira D., Steynberg J.P. , Burger J.F.W., Young D A., Brandt
E.V., Steenkamp J A.; Chem. Soc. P erkin T rans.1., (1988), 3323.
[61] Ferreira D., Steynberg J.P. , Burger J .F.W., Young D.A., Brandt
E.V., Steenkamp J A . ; ./. Chem. Soc. P erkin Trans. 1., (1988), 3331.
[62] Ferreira D., Malan J.C.S., Steenkamp J A., Steynberg J .P . ,Young
D.A.,Brandt E.V.; J. Chem. Soc. P erkin T rans.1., (1990), 209.
[63] Haslam E., Cai Ya.,Lilley T.H.; J. Chem. Soc. Chem. C om m un.,
(1990), 380.
[64] Willstatter R., Zol l inger E.H.; Justu s L ieb ig s Ann. C hem ., 412,
(1916), 195
[65] Haslam E., Mistry T.V.,Cai Ya., Lilley T . H . ; ./. Chem. Soc. Perkin
Trans. 1, (1991) , 1287.
59
[66] Asen S., S tew ar t R .N . , N or r i s K .H . ; P h y t o c h e m i s t r y , 29, (1990),
1097.
[67] Chen L.J . , H razd ina G.; P hytochem istry , 20, (1981) , 297.
[68] Brou il la rd R. , M a z z a G., Saad Z. , A lb rech t -G ary A .M ., Cheminat A.;
J. Am. Chem. Soc., I l l , (1989) , 2604 .
[69] G u s tav so n K .H . , in The C hem istry o f Tanning P rocess ;Academic
Press , N e w Y ork (1956) , 156.
[70] Loom is W .D. , Bat ta i le J., P hytochem istry ; 5, (1966) , 423
[71] Loom is W .D. ; M ethods E n zym o l ., 31, (1974) , 528.
[72] van Sumere C .V. , A lbrech t J., D e d o n d e r A., D e P o o t e r H . ,P e I.;
in The C hem istry a n d B io chem istry o f P lan t P ro te in s ,Eds. H arborne
J .B . , van Sum ere C .V . ,A cad e m ic Press , N ew Y ork , (1975) .
[73] A r tz W .E ., B ishop P .D . , D u n k e r A .K., Schanus E .G . , S w anson B.G. ;
J. Agric. F o o d C hem ., 35, (1987) , 417.
[74] Weinges K., E b e r t W., H u th w e lk e r D., M a t t a u c h H., P e rn e r J.;
L ieb ig s Ann. C hem ., 726, (1969) , 114.
[75] W einges K., M a t t a u c h H., Wilkins C. , F ro s t D .; L ieb ig s Ann. Chem.,
754, (1971) , 124.
[76] van Soes t T C . ,L ieb ig s A tm Chem, 754, (1971) , 137.
60
[77] R om eyer F .M . , Sapis J .C. , Macheix J .J . , P hytochem istry , (1985).
[78] Sapis J .C. , Macheix J .J . , C o rd o n n ie r R.E.; Am. ./. Enol. V itic ., 34,
(1983) , 157.
[79] Lea A .G .H . ; Am. ./. Etto I . V itic ., 16, (1984) , 47.
[80] S ingle ton V .L . ,T ro u s d a le E.; Am. J. Enol. V itic .,34, (1983) , 27.
[81 ]V oya tz i s Y., Glor ies Y.; C.R. Ann. Act. liech. Inst. E /io l., Bordeaux
Univers i ty, ( I 983) , 8 I .
[82] Oszmianski J. , Sapis J .C . , Macheix J .J .; J. F ood Science, 50, (1985) ,
1505.
[83] M cM urragh l . ,L o u g h re y M.J. , I l enn igen G .P .; «/. Sci. F ood A grie .,
3 4 , (1983) , 62.
[8 4 J Je rumanis J . , Fur. Brew. Conv. P roc., (1979) , 309.
[85] S ingle ton V.L . ; Am. ./. I'lnoI . V itic ., 38, (1987 ) , 69.
[86] Cheynier V., O sse C. , R igaud J.; J. F ood S c ience , 53, (1988) , 1729.
[87] Cheyn ie r V., van Huls t M . W . J . ; ./. Agrie. Pood. Cheni., 36,
(1988) ,10 .
[88] Cheynie r V., Bas ire N., R igaud J.; J. Agrie. F o o d C hem ., 37, (1989) ,
1069.
61
[89] Cheynier V., da Silva J . M . R . , ./. Agric. F o o d C hem . , 39, (1991) ,
1047.
[90] Oszmianski J., Lee C.Y , . / . Agric. Food. C hem ., 38, (1990) , 1202.
[91] Oszmianski J. , Lee C . Y . , ./. A g ric .F o o d C hem ., 39, (1991) , 1050.
[92] Oleszek W., Lee C . Y . ; Acta. Soc. Bot. P o l., 58, (1989) , 273.
[93] M a thew A.G ., P arp ia H .A .B . ; Adv. F ood R es., 19, (1971) , 75.
[94] Ponling J .D.; in F ood Enzym es, Ed. Schul tz H .W . , Avi, N ew York,
( I 9 6 0 ) .
[95] Janov i tz -KIapp A.M., R ichard F .C . , G oupy P .M . , Nicholas J . J . ; ./.
Agric. F ood C hem ., 38, (1990) , 926.
[96] S apars G .M . , Z io lkow sk i M ,A . ; 7 . F ood S c i., 52, (1987) , 1732.
[97] Embs R.J. , M arkak is P . ; ./.F ood S c i., 30, (1965 ) , 753.
[98] Tay lo r J .P . ,B ush A.H .; J. Agric. h o o d C hem ., 34(19 8 6 ) ,4 1 2
[99] Langdon T .T . ; F o o d Teclinol. ,4 I , (1987) , 64.
[1 0 0 ] J o s ly n M .A . ,P o n t in g J .D . ; Adv. F ood R es., 3, (1951 ) , 1.
[101] R obe r t s E .A .H . ; Chem. Ind ., (1959) , 995.
62
[102] Sanada M., Suzue R. , N akash im a Y., K aw ad a S.; Bi oc hi m. Biophys.
A c/a ., 261, (1972) , 258.
[103] Dudley E D., H o tchk i s s J H . , F ood B iochem ., 13, (1989) , 65.
[ 104] Richard F.C. , G o u p y P .M., N ico las J .J . , Laconibe J .M . ,Pav ia A.A.;
./. Agrie. F ood C hem ., 39, ( 1991 ), 84 I .
[105] Nicolosi G. , L am b u s ta D., P a t t i A., Piat tel li M.; Synthesis , (1993) ,
1 155.
[106] Rayle P .R. , C h a k ra b o r ty J., T hom son A.D .; Pharm acol. Biochem .
Behav., 18, (1983 ) , 473.
[107] Bonat i A., M ust ich G.; Ger. O ffen. 271 1927, (1977).
[108] K anebo Ltd. Jpn. B aleni 5 7 / IX 5X0 , (1982) .
[ 109] Kashket S.; U.S. B aleni -19064X0, (1990) .
[1 10] Kuniko K., Koichi ro O ; .Jpn. B aleni 62X1303, (1987) .
63
Experimental section:
(1) Introduction:
(2) R eag e n ts and Chemicals:
(3) A ppara tus :
(a) H P L C sys tem
(b) F reeze dry ing A p p a ra tu s
(c) N M R S p e c t ro m e te r
(d) M ass S p e c t ro m e te r
(4) Exper im en ta l P rocedu res :
(a) C o n d e n s a t io n Synthes is (D e lc o u r e t a l) [1]
(b) C ondensa t ion Synthes is (F o n k n ech te n ei a l) [2]
(c) A cé ty la t ion o f C o n d e n s a t io n P r o d u c t s [3]
(d) Enzym at ic O xida t ion o f (+ ) -C a tech in (O szmiansk i e t a l) [4]
(e) E nzym at ic O x ida t ion o f (+ ) -C a tech in (W eignes el a l) [5]
(5) Resu l t s and Discussion:
(a) Ini t ial Familia r isa t ion E xper im en t s
(i) C olum n C h ro m a to g ra p h y Vs Cel lu lose P la tes
(ii) H P L C M o n i to r in g
(b) F o n k n e c h te n e t a l Synthes is - A cé ty la t ion R eac t ions
(c) D e lc o u r el a l Synthes is - A cé ty la t ion R eac t ions
(d) Synthes is o f F r e e H y d ro x y C o m p o u n d s
(e) Enzym at ic O x ida t ion Syntheses
(6) Conc lu s ion
64
(1) INTRODUCTION:
The aim o f the work undertaken in this project was to try and synthesise
the condensed tannin dimer known as Procyanidin B3. The idea behind
this was to supply Guinness Group Research with authentic standards for
use with HPLC. In order to achieve this goal an investigation o f synthetic
procedures outlined in the l i terature was initiated. Most synthetic routes
involved the reduct ion o f (+)-taxifolin followed by the condensation o f the
reduced intermediate with (+)-catechin. Since the (+)-taxifolin could only
be obtained from commercial sources having been extracted from various
plant sources it was correspondingly expensive. Due to this expense, two
factors predominated any investigations into condensed tannin chemistry.
Firstly, all experiments had to be carried out on a small scale (largest input
o f (+)-taxifolin = 500mgs.). The second factor was to try and realise the
maximum yield o f the condensation products for the inputs used. For
these reasons it was imperative that an efficient method o f purification be
developed.
One important factor that emerged from the l i terature review was the fact
that procyanidin B3 in its free form was a fairly labile material. For this
reason it was decided to employ two strategies during the synthesis o f this
dimer. Firstly, in order to establish the worth o f each o f the methods
examined the products were isolated in their more stable peracetate
derivative form. Secondly, in order to characterise the compound in its
free form any product produced was subjected to lyopholisation in order to
perserve it. Finally, none o f the work could have been undertaken without
suitable methods o f character isation for the products. For this 400mHz ’H
N M R was employed and spectra generated compared to those persented in
l i terature studies. It was also noteworthy that this project involved
65
substantial input from analytical chemistry as HPLC methods had to be
developed that al lowed for in process assessment o f the procedures. These
analyses were scaled up to enable selective purification o f compounds by
semi-preparative HPLC.
(2) CHEMICALS & REAGENTS:
All (+)-taxifolin was purchased from Koch Light Limited and Extrasynthese
Limited and was donated for research by Guinness Group Research. The
(+)-catechin and sodium borohydride were purchased from Aldrich
Chemical Co. Limited. The enzymes, tyrosinase and horse radish
peroxidase were purchased from Sigma Chemical Co. Limited. All
solvents used were HPLC grade and were supplied by Labscan Limited.
All other reagents and solvents were ex Dublin City University Chemical
stores.
(3) APPARATUS:
(i) High Performance Liquid Chromatograph (Semi-preparative)
(a) Solvent Delivery Pump; Waters 510 Dual Head Recriprocating
Pump, capable o f delivering a flow o f 19.9mls/min.
(b) Injection Por t; Rheodyne 7125 manual injector with 20(0.1 fixed
loop (analytical) , 1ml fixed loop (semi-preparative).
(c) Detector; Shimadzu SPD-6A variable wavelength de tec tor
(190-700nm Range) used in ultraviolet at 220 & 280nm.
(d) In tegrator; Waters 746 In tegra to r (128K Memory)
(e) Fraction Collector; LKB Bromma 2000 Automatic Fractionaction
Unit (200 x 15ml tube collection facility).
66
HPLC Columns:
(a) Analyt ical co lumns used;
(i) W ate rs Rad ial C o m p re ss io n M o d u le (R C M ) 10cm x 8mm
p B o n d a p a k C18 reve rse phase pack ing , 15pm.
(ii) C h rom ex Nucleos i l C18 2 5 c m x 4 .6mm, 5pm.
(iii) p B o n d a p a k C18 3 0 cm x 3 .9mm, 10p,m.
(b) S em i -P repa ra t ive C o lum ns used;
(i) W ate r s R C M C ar t r idge Sys tem 2 0 c m x 25m m p B o n d a p a k C l 8,
15pm packing.
Freeze Drying Unit:
(i) L a b c o n c o 4.5l t . b e n c h to p lyophol i se r w i th E d w a rd s T w o S tage
V a c u u m P um p ( < 1 0 0 m B a r capabil i ty) .
Nuclear Magnetic Resonance Spectrometer:
(i) B ru k e r A C -4 0 0 - 4 0 0 m H z F T -P u l s e NMR.
Mass Spectrometer:
(i) S tand A lo n e Q u a d ro p o le M S w i th e l ec t ron ion isa t ion p ro b e (<1000
amu)
67
(4) EXPERIMENTAL PROCEDURES:
The following section outlines experimental procedures involved in the
synthesis o f procyanidin B3. In the interests o f simplicity each procedure
was examined separately with all values for the condensation reactions
listed relative to a lOOmg input o f (+)-taxifolin. For the enzyme reactions
concentra tions were more appropr ia te for their discussion.
(i) Synthesis of procyanidin B3 [Delcour Method] from (+)-taxifolin
and (+)-catechiu in a 1:5 molar ratio [1]
Approximately 500mgs o f (+)-catechin was dissolved in lOmls o f ethanol
containing approximately lOOmgs o f (+)-taxifolin in solution.
Approximately 80mgs o f sodium borohydr ide (NaBH4) was added to 2.5mls
o f ethanol (forms a suspension ra ther than solution). This was added to
the ethanol solution in a round bo t tom flask dropwise over 10 mins. under
nitrogen. Once added, 12.5mls o f water was used to dilute the solution
and the pFI was adjusted to 5.0 using 0 .15M acetic acid. The reaction was
allowed to stand at ambient temperature for 1 hour with constant agitation.
The reaction solution was diluted to 40mls with water and the resulting
phenolic compounds were extracted with 6 x volumes o f ethyl acetate.
This was evaporated to dryness by rotary evaporation having first been
dried over magnesium sulphate. The dried products were then either
freeze dried or fur ther reacted to form their perace ta te derivatives (c f later
section). The products were then subjected to purification by several
methods.
Note 1: HPLC was used for inprocess monitor ing o f the reactions.
6 8
Note 2: All weights and volumes were adjusted for larger scale
preparat ions where appropriate.
(ii) Synthesis of Procyanidin B3[Fonknechten Method] using a 1:2
molar ratio of (+)-taxifolin to (+)-catechin and optimised conditions
[2].
Approximately 1 OOmgs o f (+)-taxifolin was added to 5mls o f water at 60°C
in a round bottom flask. Approximately 50mgs o f solid N aBH 4 was added
to this slowly to avoid vigorous evolution o f hydrogen. Once added the
flask was cooled to ambient tempera ture and allowed stand for 15 mins.
with constant agitation. The flask was cooled to between -6°C and -I0°C
with the aid o f an ace tone ice bath. Approximately 50mgs o f (+)-catechin
was dissolved in 5mls o f methanol and was added to the cooling flask
before the aqueous solution froze. Once the temperature had been
achieved the solution was allowed stand for 20 to 30 mins. with constant
agitation. The react ion was quenched using 0.3mls. o f 37% HC1 solution.
The result ing products were extracted using 6 x volumes o f ethyl acetate
and isolated as outlined in the previous procedure.
Note 1: HPLC was used to moni tor the inprocess reactions.
Note 2: All weights and volumes were adjusted for larger scale
experiments where appropriate.
(iii) Acetylation of reaction products using a 1:1 ratio of pyridine and
acetic anhydride [3]
Following rotary evopora tion o f ethyl acetate the resulting residue was
re-dissolved in 2mls o f pyridine and 2mls o f acetic anhydride was added.
The flask was allowed to stand overnight at ambient with constant
agitation, The acetylated products were recovered following their
precipitation on ice. The precipitated products were continually washed
69
with cool water until no further smell o f pyridine was detected. All
products were dried under vacuum and subjected to fur ther analysis.
(iv) Enzymatic Oxidation of (+)-catecIiin using tyrosinase (niuslirooin
polyphenol oxidase) [4]
The following procedure details the work carried out for the attempted
synthesis o f procyanidin B3 by enzymatic synthesis.
(i) Prepara tion o f reagent solutions.
(a) 0 .02M Acetic Acid
A stock solution was prepared by accurately weighing 1.20gms of
acetic acid into a 1L volumetric flask which was diluted to
volume with distilled water. The pH o f this solution was adjusted
to 3.5 or 6.5 accordingly using l.ON NaOH solution. This
formed a preparat ive solution for all other solutions used.
(b) 2 .0mM Catechin Solution
Approximate ly 58 to 60mgs o f (+)-catechin was weighed into a
100ml volumetric flask. This was diluted to volume with the
0.02M stock solution.
(c) 0.5mgs/ml Tyrosinase Solution
Approximately 5.0mgs o f Tyrosinase was weighed into a 10ml
volumetric flask and was diluted to volume with stock solution.
The activity o f the enzyme was assigned at 2100 units/mg o f solid
so the final activity o f the enzyme was approximately 1050 units.
(ii) 19.5mls o f 2 .0mM (+)-catechin solution was added to a 100ml round
bot tom flask. 0.5mls o f 0.5mgs/ml tyrosinase solution was added to this at
ambient. The react ion was allowed to proceed overnight with constant
agitation. The react ion was quenched with an equal volume o f 50:50
CH3CN:(3 % v/v)HCl solution. The aqueous solution was subjected to 6 x
volume extractions o f ethyl acetate and the result ing evaporated residue
was treated in a similar manner to that already described.
70
(v) Formation of dehydrodicatechin A by enzymatic oxidation using
horse radish peroxidase. [5].
Approximately 500mgs o f (+)-catechin was dissolved in the minimum
quantity o f ace tone required to fully dissolve it,in a 100ml round bottom
flask. 20mls o f water was added at ambient with constant agitation.
lOOmgs o f horse radish peroxidase was dissolved in2.0 mis o f 0.05M
sodium citrate buffer (Sorenson buffer) , pH5.60,and was added to the
catechin solution. 1.0ml o f 0.3%v/v hydrogen peroxide was added
dropwise. The react ion was allowed to stand at ambient for 14 days with
daily additions o f the enzyme and hydrogen peroxide in the prescribed
concentrations. A yellow precipitate was observed and the solution was
gravity fil tered using Whatman N o l fi l terpaper. The wet product was
placed into a conical flask and redissolved in acetone. This was repeatedly
refluxed in the presence o f activated charcoal, in order to remove the dark
colour,and finally water was added to the hot fil trate and left fo r l2 -24
hours .
Note 1: HPLC was used to monitor the react ion throughout
71
(5) RESULTS & DISCUSSION:
(i) Initial Familiarisation Experiments.
The preliminary experiments that were completed in this field o f study
were done so in order to address some o f the problems likely to be
encountered. The first problem to be solved was the determination as to
whether or not the react ion was proceeding as described in the l iterature.
To solve this an effective method o f in process determination had to be
developed. Most o f the references examined alluded to the use o f HPLC as
the method o f analysis o f choice. It was clear that gradient elution was
used and since this facility was not available an isocratic assay that
incorporated the mobile phases used in the gradient assay was developed.
The first at tempt at synthesis was on a 100 mg input o f (+)-taxifolin
scale,and followed the procedure described by Fonknechten et a l , which
claimed a yield o f 50% for procyanidin B3. This reaction served to validate
the synthesis as well as the FIPLC assay developed. Complete development
o f the analytical assay was hampered due to the fact that only two o f the
start ing materials were available as HPLC markers, leaving the unequivocal
assignation o f the re tent ion time for procyanidin B3 open to question. The
synthesis directly followed the procedure as laid down and sampling for in
process determinations was performed on a 30 min. interval scale. The
analytical conditions were outlined as follows:
72
ANALYTICAL HPLC CONDITIONS:
Column: Chromex Nucleosil C18 25cm X 4.6mmid 10(.im packing
Mobile Phase: 90 : 10 (10%v/v)Acet ic Acid : CH,CN.
Flow Rate: 1.0 mls/min.
Detection: X = 280 nm.
Retention Times TimefMins.)
(+)-Catechin 4.90
(+)-Taxifolin 16.00
Table #1: In Process Determinat ion o f Initial Synthesis reaction
Sample/Retlime Ukl
3.33mins
Uk2
3.61 mins
Uk3
3.91 mins
Catechin
4.81mins
Uk4
5.44mins
Taxifolin
15.81 min
T= zero 3.20 15.94 21.53 30.12 4.86 24.31
T= 30min 8.44 n/d 33.04 20.08 6.18 32.22
T= 60min 5.68 n/d 32.24 19.00 6.59 34.47
T= 90min 6.54 n/d 34.77 18.06 6.02 33.99
T=120min 1.56 n/d 42.61 20.17 6.58 29.06
Uk = Unknown n/d = not detected
From this experiment it was possible to make a few observations. Firstly it
was apparent that a reaction was taking place. This was demonst ra ted by
the disappearance o f the peak with the re tent ion time o f 3 .61 mins.and the
consequent growth o f the peak at 3 .91 mins. Also the peak for catechin at
4.81 mins. was seen to decrease. Secondly it was noticed that the retention
t imes o f all the peaks were far too early and that baseline resolution had
not been achieved in all cases, leading to inaccurate % impurity index(II)
determinations.
73
It was felt therefore that an alternative assay was needed, since several
peaks o f interest were evident. Also it was necessary to investigate a
successful method for isolating and purifying the desired compounds. A
second synthesis on the same scale was s tar ted and once again monitored
by HPLC with alterations having been made to the mobile phase conditions
in an effort to solve some o f the problems highlighted. Once the react ion
had been completed and the resultant products were isolated, a method o f
purification had to be established. Guinness Group Research supplied a
method based upon preparat ive scale thin layer chromatography. The
method utilised cellulose plates and the mobile phase was made up o f 2 :1
:1 secButanol : acetic acid : water. Once run the plates were developed
with a spray solution comprised o f a 0.3 %v/v solution o f
4-dimethylaminocinnamaldehyde in 3:1 methanol : HCl(conc). Cathecin
was reported to yield a green colour, while procyanidin B3 appeared as
royal blue.
The assay conditions remained the same for the in process HPLC work with
the exception being the mobile phase. The ratios o f it's components were
altered to 92 :8 (10%v/v)Acetic acid : CH3CN, in order to delay the
retention o f all compounds and therefore aid in separaton.
i
74
Table #2: In Process Analysis using Alternative Assay Conditions
sample/RetTime
Ukl
4.24
Uk2
4.87
Uk3
5.25
Uk4
6.17
Catechin
6.80
taxifolin
23.50
T= 0 n/d n/d n/d n/d 0.84 99.16
T=1 5min 5.07 7.70 n/d 86.00 n/d 0.75
T= 0 Postcatechin Add'«
0.38 3.08 n/d 54.67 41.45 n/d
T=0 Post Acid Add'n
4.95 n/d 59.92 n/d 35.13 n/d
T=45minPost acid
7.37 n/d 63.81 n/d 28.82 n/d
T=75min
Post Acid6.30 n/d 67.14 n/d 26.55 n/d
T=100min
Post Acid5.71 n/d 69.06 n/d 25.22 n/d
Again even with the alterations that had been made to the assay the peaks
o f interest still eluted too early and were not fully baseline resolved. The
recovered products were dissolved in ethanol and spotted onto a cellulose
plate and this was developed. Two distinct bands were noticed and the
characterist ic green and royal blue colours appeared. Both bands were
scraped from the plate and the recovered products were dissolved in ethyl
acetate. This was removed by rotary evaporat ion and the collected
compounds were analysed by HPLC:
75
Table #3: Isolated Product Reaction Profiles
S a m p l e /
R e lT im e
Ukl
3.55
Uk2
4.08
Uk3
5 16
Catech
6.55
Uk4
7.57
Uk5
9.77
Uk6
12.68
Uk7
21.93
R xn P d l 0.43 2.77 47.06 35.35 9.47 0.80 2.19 1.28
1st b a n d 0.30 2.01 39.44 30.02 1 1.81 5.32 4.83 n/d
From the data generated in this experiment, two points emerged. First o f
all the separation on the cellulose plates did not appear to be sufficient to
allow purification o f the procyanidin B3. Secondly there appeared to be
several components present in the isolated material making identification o f
the B3 difficult. As a result o f these findings a third experiment was
undertaken, this t ime with double the input o f (+)-taxifolin. The idea was
to try and develop a column separat ion that would enable the fractionation
o f the components in a mobile phase that would be easy to remove. The
product demonst ra ted a severe reluctance to run on a silica column under
any mobile phase conditions, with dragging a common feature to most runs
tried. Analysis o f the recovered fractions showed silica not to be suitable
for chromatographic separation o f polyphenols. Cellulose powder was used
as the column stationary phase and a similiar series o f results were
obtained. A review o f the l i terature showed that the best results were
achieved using sephadex LH-20 as the stationary phase but this was not a
realist ic consideration in this case due to the expense involved in the
running o f sephadex columns. An alternative was provided in the form o f
semi-preparative HPLC. Although at first it would appear that this option
was also an expensive one, the flexibility involved with several different
purifications was enough to justify the initial setup costs involved.
76
INVESTIGATION INTO SCALING FACTORS INVOLVED IN SEMI
-PREPARATIVE HPLC PURIFICATIONS
In order to determine some o f the key parameters involved in
semi-preparative separat ions some o f the basic factors had to be
established. Firstly it was impor tant to ensure that both the analytical and
larger semi-preparative columns were made o f the same packing material, in
order to directly scale the inprocess assay to semi-preparative levels.
Sample loading was the first key parameter that had to be established since
correct application o f the sample led to the higher return o f purified
material. This factor was established by first o f all loading the maximum
quantity o f product onto the analytical column without overloading the
column excessively. Once this figure was calculated it could be used to
determine the best loading to be placed onto the larger column. The
following equation was used to determine the loading factor for the
semi-preparative column:
Load(pi.ep) = L o a d (sclling) X j D . U 2 x LI Where,
(D 2)2 x L2
D1 = Internal diameter o f semi-preparative column.
L I = Length o f semi prepara t ive column.
D2 = Internal diameter o f scaling column.
L2 = Length o f scaling column.
For (+)-catechin this fac tor was calculated as:
Load(prep) = 20mgs/ml X (2.5cm')2 x 10cm = 195.20 mgs/ml
(0 .8cm)2 x 10cm
This represents a scale factor o f 10 fold.
77
The second parameter that was important for the scale up was the increase
in flow that was needed to retain the retention times o f the peaks o f
interest . The following equat ion was used to predict this increase:
Qfpre,,) = Q(Sca.iug) X ( D l ) 2 where,
(D 2)2
Q = Flow(mls/min) and D = diameters as before.
The values for Catechin were predicted as follows:
Q(prep) = *0.88 mls/min X J X I O c m ) 2 = 8.59 mls/min
(0 .80cm)2
* The reason for the unusual f low is due to the pump head size on the semi
-preparative pump. The flow on the dial read 0.40mls/min but this has to
be multiplied by a fac tor o f 2.2 to calculate the actual flow.
Once these parameters were established a 500mg input scale synthesis was
undertaken to provide sufficient crude material for semi- preparative
isolation. The previous assays had shown that the peaks o f interest eluted
too early and it was there fore decided to use 10% v/v acetic acid as the
mobile phase. This was a highly unusual mobile phase and needed
alteration before it could be used. The stationery phase o f C l 8 HPLC
column is stable be tween a pH range o f 2.0-8.0 and for this reason the
acetic acid solution had to be adjusted to a pH grea ter than 2.0. One good
reason for using this mobile phase was the ease with which the products
could be extracted into an organic solvent. The in-process analysis for this
experiment was summarised as follows:
78
Table #4: In P ro c e s s Analys is o f 500m g Inpu t M ate r ia l
Sample/ Ret Time System
3.61
U k l
4.10
Uk2
4.56
Uk3
5.80
Catechin
6.20
Uk4
7.29
TO N a B H 4
Add'n
n/d n/d n/d 99.08 0.91 n/d
TO Cat
Add'n
0.37 n/d n/d 77.21 22.42 n/d
T=20 post
Cat Add'n
0.69 3.73 61.80 n/d 32.50 1.29
Note 1: Very poor chromatography with no baseline resolution.
Note 2: The in-process analysis was run using 92:8 10% v/v acetic acid:
CH3CN.
The next problem addressed was the type of solvent that should be used to
dissolve the compound. The use o f the above mobile phase was limited as
the product was not fully soluble. The compound was readily soluble in
methanol but injecting neat methanol using a rheodyne injector has been
known to cause peak broadening making its use non-viable. The semi
-preparative assay conditions are summarised as follows:
Column: [xBondapak C l 8 10cm x 25mm id RCM cartridge.
Mobile Phase: 10% v/v acetic acid solution.
Flow Rate. 5mls per min.
Detection: X = 280nm.
Reten tion Times: Time (mins.) Based on analytical column
B3 Suspect: 7.14
Catechin: 13.56
79
Table # 5: Tn-process Analysis p r io r to pur if icat ion.
Sample/ Ret Time U k l Uk2 Uk3 Catechin
Product 1.22 58.30 7.42 33.05
The product was made up to a final concentra tion o f 50 mgs/ml as it was
felt that a concentra tion below the optimum loading would give baseline
resolution for separation purposes. Five 1ml injections were passed down
the column and fractions collected manually into test tubes. Analysis o f
these fractions by analytical HPLC showed appreciable improvement in the
% II o f the B3 suspect. Due to this a more rigid collection routine was
established, whereby the peak o f interest was shaved to give far more
fractions o f less volume. Once completed the purified fractions were
analysed by HPLC (App. A C hrom #l) . The isolated product was found to
have a % HI purity o f 87.02% with 10.75% o f an unknown that eluted after
the B3 suspect and before catechin. No catechin was detected, however,
spiking o f the sample showed catechin to co-elute or elute close to the
unknown peak. The reason for the presence o f this unknown/ catechin
may have been due to one o f three reasons. Firstly, the procyanidin B3
may not have been stable in this particular mobile phase and may have
degraded to the later eluting peak. Secondly, the separa tion on the column
was insufficient to yield material o f the required purity for fur ther NMR
studies. Finally, the isolated product may also break down during the
organic extraction procedure used to isolate it.
Even though the isolated material was only assumed to have a purity
approaching 87.0% it was decided to analyse it by 400mHz NMR. Before
any determinations o f the product were carried out it was decided to fully
characterise (+)-catechin by N M R as some o f the spectral information
obtained by this may be common to the product isolated. The authentic
(+)-catechin standard was examined using 'H NMR, 13C NMR, 2-D COSY
80
and C-H correlation studies. This data was invaluable as catechin formed
the monomeric backbone o f the desired dimer and other higher oliogmer
compounds.
Characterisation of (+)-catechin by 400inHz NMR:
Table # 6: (i) 'H NMR spectrum (App. B NMR# I)
Proton No: Chemical Shift
(ppms)
Signal Type Splitting
(J FIz)
3 X H aromatic 6.70- 6.60 doub7doub ZJ = 12.07
1 X H 5.85 singlet ___
1 X H 5.76 singlet —
1 X H 4.50 doublet 6.54
I X H 3.95 multiplet —
1 X H 2.70 - 2.60 doub,doub I J = 10.93
1 X H 2.25 multiplet —
From the integrat ion nine protons were distinguished for (+)-catechin.
Table # 7. (ii) 13C NMR spectrum (App. B NMR#2)
Saniple/öppmCl C2 C3 C4 C5 C6 C7 C8 C9 CIOCll C12 C13 C14 C15Catechin 28.468.782.895.596.3100.8115.2 116 120132.1146.1146.1156.8 157.5157.7
From the chemical shift values obtained 15 carbons were found in
(+)-catechin having the following arrangement:
(a) 7 Carbons with no H attached (Aromatic)
(b) 5 Carbons with FI a ttached (Aromatic)
(c) 3 Aliphatic Carbons
81
(iii) 2-D COSY NMR spectrum (App.B NMR#3)
The advantage o f this spectrum was that a two dimensional plot o f the
compound was obtained which aided in the assignation o f which protons
were coupled to each other. The first plot is a contour layout o f the
spectrum along the diagnal o f identical axes. The second plot spatially
related components along either side o f the diagnol and once these
components could be linked by a square or a series o f squares the
components were said to be related. The related components were
summarised as follows:
Table #8: Chemical Shifts o f Related Peaks
C h e m ic a l
S h i f t ( p p m s )
1 St
Peak
2nd
Peak
3rd
Peak
2.50 2.80 4.15 —
2.80 2.50 4.15 —
4.15 2.50 2.80 4.75
4.75 4.75 — —
6.05 — ___ —
6.85 — — —
(iv) C-H correlation spectrum (App.B NMR#4)
(a) Dept 45 for project ion unto C-H correlation.
A specialised spectrum called a Dept 45 involves scanning f o r 13C carbon
with hydrogens a ttached which form posit ive peaks if they are even and
positive peaks if they are odd (Positive = Vertical) . Quaternary carbons
are excluded as are solvent peaks. From this spectrum it was apparent that
there were 8 carbons with hydrogens attached and that they have the
following chemical shifts:
82
I
Table # 9: D ep t 45 l3C Chem ical Shif ts fo r C -H P ro jec t ion .
Sample/ôppms C l C2 C3 C4 C5 C6 C7 C8
Catechin 28.4 68.7 82.8 95.5 96.3 100.8 115.2 116
(b) 'H NMR spectrum for projection unto C-H correlation,
Table # 1 0 : 'H NMR data for Projection Spectrum
Chemical Shift
(ppms)
Signal Type Splitting
(Hz)
6.71 singlet —
6.65 - 6.58 doub,doub £ J = 12.00
5.80 singlet ___
5.73 singlet —
4.50 - 4.35 doublet 5.68
3.88 - 3.82 multiplet —
2.75 - 2.69 doub,doub EJ - 10.10
2.40 - 2.35 multiplet —
83
(c) C-H correlation data.
Table # I 1:C-H Correla tion N M R data for Projection Spectrum
Carbon No/ôppms 'H ^ ô p p m s) lH 2(ôppms)
C l 28.4 2.40 - 2.35 2.75 - 2.69
C2 68.7 3.88 - 3.82 ----
C3 82.8 4.50 - 4.35 -----
C4 95.5 5.73 -----
C5 96.3 5.80 ----
C6 100.3 6.65 - 6.58 -----
C7 115.2 6.65 - 6.58 -----
C8 116.0 6.71 -----
Since the s tructure for (+)-catechin was already known these spect ra served
to confirm its assigned s tructure as follows:
8 '
84
By contrast t h e 'H NMR spectrum for the suspect procyanidin B3 product
was very poor. There were a large number o f interferring peaks that
further questioned the purity o f the isolated material (App.B NMR # 5).
Two courses o f action were adopted following this result. The first was
to try and alter the mobile phase for the analytical HPLC assay thereby
allowing improved analysis involving photodiode array detection. The
second was to submit a fraction o f the sample to Guinness Group Research
for analysis using a HPLC assay for which the retention time for B3 was
already assigned.
Investigation into alternative mobile phase composition.
Since the purity o f the isolated compound was questionable a range o f
alternative mobile phases were examined in order to try and afford a more
accurate purity determination. The mobile phase composi t ions examined
were summarised as follows:
(i) 85:15 10% v/v acetic acid: CH3CN
(ii) 85:15 H 20 : CH3CN
(iii) 85:15 0.5% v/v H 3P 0 4 : CH3CN
(iv) 85:15 0 .5 % v/v H 3P 0 4 : CH3CN (pH adjusted to 2.0)
(v) 80:20 CH3CN: H 2 0 (pH<3.0)
(vi) 95:5 10% v/v acetic acid: CH3CN
The first o f these when run gave a re tent ion time for (+)-catechin o f 8 .81
inins. However, the peak obtained was very broad and showed signs of
tailing. The second mobile phase showed both fronting and tailing
characterist ics and as a result was discounted. The third also gave a broad
peak. The adjustment o f pH for this mobile phase gave a well defined
peak with a retention t ime o f 1 0.6 mins. at a flow rate o f J .0 ml. per min.
Finally in order to investigate whether any late eluting compounds were
present the mobile phase was altered to (v) above with catechin eluting at
3.30 mins. There was however no evidence o f any late eluting products
which further served to confirm that procyanidin B3 had an earlier
85
retention time than catechin. Finally, the mobile phase (vi) was used to
examine the two compounds. Although this did not afford the best
separation between the two compounds it did demonst rate the sharpest
peaks. Catechin had a re tention t ime o f 4.65 mins while the procyanidin
B3 suspect eluted at 4.10 mins.
Photodiode Array Detection (PDA):
PDA detect ion was used to try and establish the integrity o f the suspect
compound. The mobile phase used was as (vi) above but the flow rate was
lowered to 0.5 mis. per min. This gave (+)-catechin a retention time of
11.56 mins. The contour plot for (+)-catechin showed a maximum
between 280-284 nm indicating that the choice o f wavelength used was
appropriate. Analysis o f (+)-catechin yielded a single peak.
(App.AChrorn#2). The retention t ime for the suspect compound was
found to be 8.75 mins, however, detailed analysis o f this compound was not
possible due to insufficient sample quantity.
Guinness Group Research Purity Determination:
As a result o f the poor NMR spectrum obtained it was decided to consult
the Analytical Department o f Guinness Group Research with the aim o f
analysing the isolated fraction on a system upon which the retention time o f
procyanidin B3 was known. They examined the compound using the
following HPLC conditions:
Column: Reverse Phase C18 25cm x 4.6 mm id, 5|.im packing.
Mobile Phase: 80:20 0.25% v/v FI3P 0 4 : methanol.
Flow Rate: 1.0 ml per min.
Detection: X = 220nm.
86
Retention Times:
Unknown# 1
Procyanidin B3
Time ( M i ns . )
(+)-Catechin
Unknown#2
5.92
7.75
9.50
12.70
Analysis by this Group showed that the product had a % HI o f only 66.0%
with 14% o f unknown #1, 18% catechin and 2% o f unknown #2. It would
appear that the use o f a lower wavelength was useful for observing peaks
other than B3 and (+)-catechin. (App.A Chrom#3)
Confirmation o f the products lower purity prompted a few changes in the
method o f analysis. It was felt that this was the key area since purification
o f the compound was only feasible by semi- preparative HPLC. The radial
compress ion module had a I5(.un packing and this was claimed to be as
effecient as a lO^im packing in a stainless steel column. A direct
comparison between the car tridge system and the more conventional steel
column showed much better chromatography for the lat ter and thus the
car tridge system was discarded for analytical determinations.
Having defined many o f the parameters involved in both analytical and
semi-preparative work a 500mg input o f (+)-taxifolin was undertaken
following the method o f Fonknechten el a l and the final recovered weight
was calculated at 680mgs.
Theoretical Maximum Recovery (100% conversion + 100% recovery)
500 m g s. (+)- tax ifolin x 579 .3mol wt. B2 = 957.8mgs.
302.4mol wt. (+)-taxifolin
Assume only 50% yield implies theoretical max. = 478.9mgs.
Total recovered compound = 680mgs.
87
% B3 (by HPLC) = 33.45 %= 680.00mgs x 33.45% = 227.50mgs
1 00%
Therefore % recovery o f B3 = 2 2 7 .50mgs x 100% = 47.50%
478.90mgs
In theory therefore the implied 50% yield attr ibuted to this synthesis held.
Table #12: HPLC Profile o f isolated product
Sample/Ret
Time
ukl
5.01
uk2
6.23
uk3
7.19
uk4
8.20
uk5
9.86
uk6
11.01
Catn
14.40
uk7
16.03
uk8
18.24
uk9
23.52
Product 0.98 1.53 1.01 33.4 3.72 11.08 33.68 11.62 0.31 0.83
The isolated product was split into two fractions and two attempts at its
purification by semi-preparative HPLC were made. The first used a
concentrat ion o f lOmgs/ml and the time window be tween 17.0 and
18.6mins. for the fraction collection. The isolated material only had a %
HI o f 67.3% with the o ther unknown impurities making up the difference.
The second attempt used a more concentra ted solution at 12mgs/ml. This
lead to the recovery o f a peak with a retention time o f 11.5mins with a %
HI purity grea te r than 90% for the collected fractions. However, when
this was analysed by 400mHz NMR it was apparent that the material was
impure and that a large quanti ty o f catechin was still present.
Even though the recovered product was not pure there were some
distinguishable features in the 'H NMR spectrum. The presence o f
aromatic hydrogens at a chemical shift greater than 8.0ppms was
unexpected and sugges ted that the peak collected was not that for
procyanidin B3.
88
In l i te ra ture |6J it was s ta ted that procyan id in B3 in its free form was a
labile material and this meant tha t the iso la ted p roduc t was probably
suscep tib le to d e g ra d a t io n dur ing iso lat ion. Since no firm ev idence for the
p roduc t ion o f procyan id in 133 had been ob ta ined us ing the synthesis
c ond i t ions as laid do w n by F o n k n e c h te n e t a / it was dec ided to regress to
the original (and m o re r igo rous ) synthe t ic p ro c e d u re descr ibed by D e lco u r
e t a l .
89
Synthesis of Procyanidin B3 from (+)-taxifoIin and (+)-catechin in a
1:5 molar ratio.
Following the procedure detailed in the experimental section a reaction
using 200mgs input o f (+)-taxifolin was undertaken. The more stringent
conditions involved in this synthesis included the use o f ambient
temperature and a nitrogen atmosphere , in order to try and control the
initial reduction and subsequent condensa tion reaction. The reaction
appeared to have proceeded as described with all steps being monitored by
HPLC. The in-process details were summarised as follows:
Table #13: In-process analysis.
Sample/Ret
Time
ukl
8.90
uk2
10.80
uk3
14.00
Catechin
16.10
Taxifolin
75.00
TO preNaBH,, n/d n/d n/d 99.24 n/a*
TOpostNaBH,, 0.26 1.24 12.80 81.50 n/a
TO post pH 0.11 n/d 12.16 87.08 n/d**
T = l H r hold 0.12 1.35 13.74 84.04 n/d
T=2Hr hold 0.10 1.29 13.08 83.10 n/d
* N/A = Not Analysed ; N/D = N ot Detected
From the above analysis it was observed that the procyanidin B3 had
formed in a 13% yield. This was less than the proposed 20% yield claimed
but this may be explained due to the small scale o f this experiment. Once
the reaction was complete the products were extracted in the same fashion
as the previous experiments, using ethyl acetate. The isolated product had
the following react ion profile;
90
Table# 14: Isolated Product Reaction Profile.
Sample/Ret
Time
ukl
3.25
uk2
5.02
uk3
8.24
uk4
9.97
uk5
10.71
uk6
13.30
Cat
14.40
uk7
15.90
uk8
17.80
uk9
19.50
Product 0.18 0.14 21.4 1.37 0.94 0.10 70.27 4.45 0.58 0.33
This material was purified using semi-preparative HPLC and the peak o f
interest was frac tionated into forty tubes. Analysis o f the purified fractions
showed that the first 9 fractions contained the desired compound with a
%HI purity o f 95.0% or bette r (App.A Chrom.#4). The combined fractions
were extracted using ethyl acetate in the usual manner. The result ing
residue was analysed by 400mHz NMR. The spectrum showed interference
from ethyl acetate in an impor tant part o f the spectrum. The spectrum
showed resonances, previously observed in l i terature assignations o f
procyanidin B3. One observat ion that was noteworthy was the fact that the
spectrum exhibited signs o f resonance broadening and this was consistent
with spectra reported in l i terature reviews and which were att r ibuted to
conformational isomerisation. Some o f the key features o f this spectrum
were compared to l i terature assignations:
Table# 14: Spectral details o f Isolated Product (Delcour Methodology) .
Pro ton Type Chemical Shift
(pptns)
Signal Type Splitting
(Hz)
6 X Ar-H[n/ErijigJ 7.00 - 6.56 multiplet -----
6.16 singlet ----
8 - H [ a | , 6 -H[a) 6.05 singlet -----
3 -H [c] 5.84 doub,doub 3"oo00JJ
91
The following diagram affords an assignment o f the peaks defined by the
above table:
Procyanidin B3 Model + Numbering
HO
1011
B
OH
8
A
O
C
12
OH
-\ 5 / 4
14
OHHO
1k -
F
L
OH
OH
Table# 15: Spectral details o f Isolated product:
P ro ton Type Chemical Shift
(ppms)
Signal Type Splitting
(Hz)
7.20 singlet -----
6 X Ar- I I [n/EHng| 7.00 - 6.65 multiplet -----
6-H [a] ; 8 -H[a1 6.50 doublet 2.57
4-H lc] 6.29 doublet 7.70
6.20 singlet -----
6.05 singlet -----
6.02 singlet -----
5.95 singlet -----
5.85 doub,doub 2 J = 17.78
5.65 singlet -----
92
Even though the recovered product showed evidence for the presence o f
procyanidin B3 it was not o f sufficient quality to allow a complete
identification. Also, there was insufficient quantit ies to allow for a second
purification. A separa te attempt at isolation was undertaken using the
remaining quanti ty o f this material. This used a more concentra ted input
which it was felt might lead to a grea ter yield o f purified procyanidin B3.
All the conditions were kept the same for both at tempts and once again
only the material from the first 9 fractions was found to be o f sufficient
purity to attempt isolation. A rough at tempt at the determination o f the
concentra tion o f the product not extracted by the ethyl acetate showed that
almost 20% o f the recovered material failed to be extracted and hence was
lost. The material was analysed by 400mHz NMR but once again was not
pure enough to allow identification.
One final experiment was undertaken and again an attempt at purification
using semi-preparative HPLC undertaken. The concentra tion o f the
solution loaded onto the column was 100.00 mgs/ml (approximating
optimum loading). Once isolated the product was freeze dried and
analysed by HPLC. This showed at least 16% (+)-catechin was still
present and this was not unexpected as catechin was present at five times
the concentra tion o f the (+)-taxifolin start ing material. Also it was felt
that at higher loadings loss o f resolution would lead to catechin
contamination. The resulting product was purified a second t ime and
although HPLC analysis o f the collected fractions appeared pure catechin
was observed to be present upon isolation. The main problem with this
second purification was the low yield o f the recovered material. By contrast
the volume o f the collected fractions remain the same making quali tative
analysis by HPLC impossible. Analysis o f the isolated product showed
catechin to be present at 30% II. This was not thought to be from impure
isolation but ra ther from the de-composi t ion o f the product to yield
(+)-catechin.
93
It was felt that this may have occurred during the final extraction into ethyl
acetate (App.A Chrom#5). This called into question the stability of
procyanidin B3 in its free form and lead to analysis o f this compound in its
acetylated derivative form.
Acetylation of Condensation Products:
Most analytical studies carried out 011 condensed tannins were done so
using acetylated or methylated derivatives o f these compounds as this
greatly aided their stability. The lack o f success in isolating procyanidin
B3 in its free form prompted the use o f such dei ivatisation techniques. In
order to employ this procedure some issues again had to be resolved, most
notably, the development o f a suitable HPLC assay that would allow
separation o f the acetylated compounds. In order to prepare a peak
marker for assay development a repeat o f the Delcour et a l synthesis was
undertaken with the difference that once extracted from ethyl acetate the
products were re-dissolved in a 1:1 equivalent o f pyridine:acetic anhydride.
This was allowed to stand at ambient tempera ture overnight with constant
agitation. The products were recovered by their precipitation onto ice and
the recovered material was repeatedly washed with chilled water until no
smell o f pyridine was apparent.
Several different mobile phase composit ions were tried in which both
methanol and acetonitri le were used in conjunction with w ater until a
suitable assay was developed. The column used was that used in the
previous experiments but the mobile phase was altered to 50:50CH3CN:H20
(ApH = 3.0). This gave retention times for the acetylated products that
approximated those o f the products in their free form in the previous assay
All other parameters remained unchanged. One major advantage to this
assay was the lack o f buffers or organic modifiers which therefore allowed
the assay be used directly for semi-preparative HPLC work.
94
The composit ion o f the acetylated compounds showed two main peaks,
thought to be (+)-catechin and the procyanidin B3 suspect. In order to
confirm the order o f elution it was first necessary to acetylate (+)-catechin
as a peak marker. This also proved useful for analysis by 400mHz N M R
and mass spectroscopy as the resonances and ions o f this compound would
most likely be found in the spectra generated for the product. The HPLC
profile for acetylated catechin was summarised as follows.
Table #16: HPLC Profile for Acetylated Catechin
Sample/Ret
Time
u k l
7.06
uk2
7.99
uk3
8.34
uk4
8.71
Catn
13.40
uk5
19.12
uk6
21.87
Cat-oAc 0.15 0.41 0.27 0.38 98.40 0.10 0.13
The mass spectrum (App.2 spec #1) for acetylated Catechin gave a
molecular ion (M/e) o f 500, consistent with the mass expected for
acetylated catechin. The *H N M R for the compound was summarised as
follows: (App.B N M R#6)
Table #17: 400mHz N M R Spectral Details:
Pro ton Type Chemical Shift
(ppms)
Signal Type Splitting
(Hz)
3 X Ar-H[u| 7.30 - 7.15 multiplet ----
1 X Ar-H[A] 6.65 singlet -----
1 X Ar-H [A] 6.55 singlet -----
5.25 multiplet -----
3 -H [C] 5.15 doublet 3.70
2.85 doub,doub I J = 18.78
4-H îbj 2.65 doub,doub SJ = 17.45
OAc 2.25 - 2.00 ----- ----
95
The 13C NMR spectrum was allocated as follows. (App. B NMR#7)
Table #18:
CIIEMSHIFT C l C2 C3 C4 C5 C6 C7 C8 C9 CIO C l l C12 C13
Catn 20.61 20.64 20.77 20.95 21.08 23.9 68.27 77.64107.68108.78 110.2121.76 123.7ii C14 C15 C16 C17 C18 C19 C20 C21 C22 C23 C24 —- —it 124.4136.13142.07 142.1149.41149.85154.36168.08168.39168.98170.15 — . . .
From this spec trum it was observed that there were 5 carbons bound to
oxygens, there were 4 aromatic carbons bound to oxygens, 4 aromatic
carbons, 1 aliphatic carbon bound to oxygen and 6 carbons assigned for the
C H 3 groups o f the acetylated compound. To augment these findings a
D EPT 45 projection for the carbons was obtained. As suspected from the
13C trace there were 5 peaks between 21.0-20.5 ppms that corresponded to
the CH3 groups. There was a CH group at 25.0 ppms and fur ther CH, CH2
groups at 124.35, 123.70, 121.70, 108.60, 107.60, 77.60 and 68.20ppms.
These peaks were the carbons bound to oxygens and aromatic carbons
bound to hydrogen that were expected In all 13 out o f the 25 carbons
were bound to hydrogens with 5 bound to oxygens leaving a total o f 7
bound to other carbons.
96
Finally, a C-FT corre la tion was used to fur ther determine the posit ions at
which carbon was bound to hydrogen. The following table details those
couplings: (App. B NMR#8).
Table #19: Summary o f C-H Corre la tion Data
13C / ‘H H l ( 2 . 15) H 2 (2 9-2 5 5 ) H3(5.05) H4(5 .15) H5(6.40) H6(6.55) H7(7.10)
C l( 2 0 .6-21.0) X
C 2 ( 2 3 . 9 ) X
C3 ( 6 8 . 3 ) X
C 4 (77.6) X
C5(107.6) X
C6(108.7) X
C7(121.7) X
C8(123.7) X
C9(124.3) X
From the findings observed from the acetylat ion o f (+)-catechin it appeared
that this method was both effective and successful and that the assay
developed did in fact work in the separation o f the acetylated compounds.
An experiment using 400 mgs input (+)-taxifolin was undertaken to provide
acetylated material for purification. Once the material was isolated post
acetylation it was examined by HPLC and pyridine was observed to be still
present. The material was further washed and dried under vacuum but the
interferrence on HPLC was seen to remain causing doubt as to whether this
was in fact pyridine. The material was subjected to a double purification
by semi-preparative HPLC. However, following the two separations
insufficient quanti t ies o f isolated material were recovered. As a result o f
this it was decided to repeat the acetylation using the method of
Fonknechten et a l to yield the crude products ingreater yield.
97
The crude product post acétylation had a broad peak at a retention time o f
8.4 mins at 36 .13%II and a peak at 12.96 mins at 56 .35%II. Further
analysis o f this compound showed a late eluting peak at 25 mins and this
was assumed to be procyanidin B3. It was therefore assumed that the first
broad peak was most likely residual pyridine. Once again the product was
dried using vacuum to try and remove this peak.
The re-dried material was analysed by HPLC which resulted in the
observat ion o f 3 main peaks with retention times at 12.80, 30.50, and 59.90
mins respectively (App.A Chrom#6). It was decided to collect the later
two peaks since the first peak was identified as acetylated catechin by
spiking with this standard. A summary o f the reaction profile was as
follows:
Table #20: React ion Profile for Isolated Acetylated Compound
Sample/Ret
Time
uk l
7.65
uk2
10.5
uk3
11.8
uk4
12.4
Cat 'n
12.8
uk5
17.8
uk6
28.1
B3?
30.5
uk7
33.4
uk8
36.9
uk9
54.0
uklO
59.9
Acet 'd Pdt 0.87 0.08 1.92 0.69 24.7 0.85 0.4 24.3 2.56 3.18 3.3 23.9
This material was purified using an initial concentra tion o f 65.60 mgs/ml.
The peak at 30.50 mins was fractionated into 40 tubes in order to improve
its purification. Analysis o f the tubes 20 through 40 showed a good
quality for the desired peak and these fractions were combined giving a
compound with 83 .50%HI purity prior to extraction. The late eluting peak
was collected as one fraction since no other peaks were observed to
interfere with it. This had a % HI purity g rea te r than 90%. Both o f these
compounds were isolated in the usual manner and both were subjected to
analysis by 400mHz NMR.
98
NMR Analysis of 30.54iniu. peak:
(i) ’H NMR Spectra Data.
The following s t ructure was used to detail the proton posit ions o f
acetylated B3.
14PROCYANLDIN B3
cAO » 112
p A c13 OAc
13
r-O- \ 10
11
5 ^ An ■*
O1 114 iu ^o A C
/ OcA
cAO
/ II3’ 'OAcII4'ax H4'i(|
OAc
From the values reported by Fonknechten et «/( table 21) for acetylated B3
a table for the recovered material was compiled with all key pro ton
resonances listed:
Table #21: Fonknechten et a / Acetylated B3.
Pro ton Type Chemical Shift
(ppms)
Signal Type Splitting
(Hz)
H-Ar 7.40 - 6.45 multiplet -----
3 -H [c] 5.06 triplet -----
3 ’-H (f] ; 2 ' -H lF1 4.95 multiplet -----
2 -H [c] 4.78 doublet 9.80
4"H[cj 4.50 doublet 9.80
4'-Heq 2.96 doub,doub I J = 16.5
4'-Hax 2.66 doub,doub EJ = 16.5
99
Note 1: All other aliphatic protons were attr ibuted to acetyl groups.
N ote 2: Resolution o f spectrum suggested 80mHz N M R was used although
this was not stated.
Table #22: Recovered peak 'H NMR data (App B NMR#9).
Pro ton Type Chemical Shift
(ppms)
Signal Type Splitting
(Hz)
H- Ar 7.20 - 6.45 multiplet -----
3-H[CJ 5.55 triplet ----
3'-Hm ; 2 '-H[f] 4.90 multiplet -----
2 -H [CJ 4.65 doublet 11.40
4 -H [c, 4.40 doublet 10.00
4'-Heq 2.85 doub,doub EJ = 17.28
4'-Hax 2,60 doub,doub I J = 21.88
Note 3: Again all alaphatic protons were at tr ibuted to acetyl groups
Analysis and confirmation o f the isolated s tructure was fur ther supported
with the aid o f a COSY spectrum. Here all the J coupled pro tons were
identified giving an indication as to which protons were coupled to each
other. (App. B NMR#10). Here some o f the bonding pat terns o f the
aromatic region were clearly defined. The doublet at 4.40ppms was
coupled to the triplet at 5.55ppms and also the doublet at 4.65ppms. This
tended to confirm the posit ions for pro tons as proposed by Fonknechten et
a l were the 3 -H[c] was splitting the 4 -H [c] into a doublet and the 2 -H [c] into a
doublet with each o f these splitting the other into a doublet . Likewise
both o f these protons split the 3 -H lc] into a triplet.
1 0 0
From comparison o f the isolated product to the product presented by
Fonknechten et a l it was apparent that their method did, in fact, produce
procyanidin B3. The peak at 30.54 mins was confirmed as acetylated B3
thereby confirming the synthesis worked However, two o f the claims
made by this group may not be correct . Firstly, it would appear that the
recovery o f procyanidin B3 in a 50% yield was not possible since a third
late eluting peak with equal % area by HPLC (assuming similar response
factor) was observed. Secondly, it was stated that this method gave no
evidence for the production o f the tr imer procyanidin C2. However, the
following NMR data detailing the spectral findings for the compound
isolated at a re tention time o f 60 mins would appear to confirm the
production o f this trimer. The first table summarised the 'H N M R data for
procyanidin C2 as proposed by Delcour et al, while the second presented
the spectral data o f the isolated material.
Table #24: ’H NMR data for procyanidin C2.
Proton Type Chemical Shift
(pptns)
Signal Type Splitting
(Hz)
9 X H-ArlB/E/H] 7.33 - 6.44 multiplet -----
6-H [d] 6.65 singlet -----
6-H [gj 6.61 singlet -----
6.53 doublet 2.50
6 -H[A] 6.20 doublet 2.30
3-H|Coi]7) 5.57 doub,doub SJ = 19.0
3_H (ForC] 5.45 doub,doub EJ = 18.5
2 -H m ; 3 -Hm 5.23 singlet(Broad) -----
2 -H (CorF] 4.73 doublet 10.25
2 " H jFoiC] 4.61 doublet 10.00
4 " H fCorF] 4.56 doublet 8.25
4"H[ForC1 4.13 doublet 9.00
1 0 1
N o te 1: All o th e r p ro to n s w ere al iphat ic and acetyl related.
Table#25: ’H N M R data for 60.00min isolated product (App. B N M R # 1 1).
Proton Type Chemical Shift
(ppms)
Signal Type Splitting
(Hz)
9 X H -A r ll)/Ii/n] 7.25 - 6.45 multiplet -----
6"H[D] 6.65 singlet -----
6 -H [G] 6.61 singlet -----
8-H [a] 6.53 doublet -----
6 ' H [a, 6.20 doublet -----
3“H lCorFJ 5.55 doub,doub EJ = 19.50
3 -H [ForC) 5.45 doub,doub EJ = 19.00
; 3 -H (I1 5.23 singlet(Broad) -----
^ -H|CorF] 4.73 doublet -----
2_H[irorC] 4.61 doublet -----
4 -H [0orF] 4.56 doublet -----
4"H[ForC] 4.13 doub,doub -----
As can be seen both the tables are closely identical which suggests that the
late eluting peak is procyanidin C2. Again a COSY spectrum was run and
the J coupled pro tons showed which pro tons were inter- related. Here the
3 doublets between 4 .73-4 .56 ppms were linked to the 2 doublet o f
doublets between 5.55-5.45 ppms (App. B NMR#12). Again 1 proton split
the other into a double t while the doublets themselves were observed to
split 3 -H[CorF] into a doublet o f doublets while 3 -H[c orF] split 2 -H (I],[c] or [F]
into a doublet.
102
Finally, what this work showed was firstly the synthesis o f acetylated
procyanidin B3 and procyanidinC2 did occur but, secondly, that these were
present in approximately 30% yields and this did not represent a major
improvement on the proposed synthesis o f Delcour et a l which was far
more rigorous. The acetylation o f these products proved useful for their
identification but was o f no use in the long term as the aim o f this work
was to isolate procyanidin B3 in its free form. Since the gain in % terms
o f the method proposed by Fonknechten el a l was little be tter than the
original method o f Delcour et a l it was decided that a t tempts at isolating
procyanidin B3 in its free form would be bet ter suited to the more rigorous
reaction o f Delcour et al. Also it was felt that with this better control o f
the synthesis, procyanidin C2 production was less likely thereby making
the separation o f procyanidin B3 by semi-preparative HPLC an easier
prospect.
103
Attempted isolation of procyanidin B3 in its free form:
In order to aid the identification o f procyanidin B3 in its free form an
authentic sample o f procyanidin B3 was supplied by Guinness Group
Research as a methanol solution stored at -20°C. This standard was
extracted from barley and was analysed by both HPLC (App. A Chrom#7)
and 400mHz NMR. The standard was evaporated from the methanol and
reconst i tuted in deuterated acetone to obtain a 'H N M R (App.2 NMR#13).
Some o f the solution was also diluted with mobile phase and was used as a
HPLC peak marker.
Table #26. 'H N M R characterist ics o f authentic B3.
Proton Type Chemical Shift
(ppms)
Splitting
(Hz)
H- Ar 7.38 - 6.65 multiplet
8 'H|A| 6.50 doublet
6.48 doublet
3 -H[CJ 6.28 doublet
3 -H [r] 6.26 doublet
2"H[Fj 6.20 singlet
2-H.c, 6.09 singlet
4-H lcj 5.94 doublet
4 -H m 5.85 multiplet
4.72 doublet
4.62 - 4.35 multiplet
4.92 doublet
104
The HPLC profile was summarised as follows
Table #27: In process HPLC Profile.
Sample/Ret
Time
ukl
0.04
uk2
2.40
uk3
2.83
uk4
3.76
uk5
4.14
uk6
4.84
uk7
5.72
uk8
7.68
B3?
9.10
uk9
10.86
B3 Std 0,02 0.87 1.2 0.07 0.42 0.24 1.76 0.33 94.5 0.54
This represented the first unambigous identification for the retention time
o f free procyanidin B3. Having established this retention time a large
scale (500 mg input (+)-taxifolin) reaction was undertaken, initially
following the procedure o f Fonknechten et al. The synthesis appeared to
go to completion in the usual manner and the crude material was isolated as
all previous reactions had been. This product was subjected to purification
by semi-preparative HPLC and analysis o f the collected fraction indicated
the presence o f a peak that shared the same retention time as procyanidin
B3 with a purity grea te r than 91 .00%HI (App. A Chrom #8). The profile
for the combined fractions was summarised as follows:
Table #27: Combined Frac tion Reaction Profile
Sample/Ret
Time
ukl
2.41
uk2
5.60
uk3
8.05
B3?
9.10
uk4
10.46
Catechin
14.83
uk5
17.04
Frac'ns 1.68 0.4 0.53 91.01 5.93 0.23 n/d
This material was extracted into ethyl acetate and evaporated to dryness.
It was immediately reconsti tu ted in CDC13 and analysed by 400mHz NMR.
The collected spectrum showed that this material was not o f sufficient
purity to yield definite p ro o f as to the presence o f procyanidin B3. Since
the purity appeared to be good by HPLC two questions again arose. The
first concern was that an impurity may have been present that was not
apparent on HPLC. The second concern was one already expressed, which
105
was the potential lability o f the compound. It was likely that the isolated
compound had degraded during its extraction. Some small evidence that
procyanidin B3 was present was presented by some key chemical shift
values as follows:
Table #28: Key Procyanidin B3 Chemical Shifts.
P ro ton No: Chemical Shift
(ppms)
Signal Type
(Isolated Product)
Signal Type
(Procyanidin B3)
1 6.40 doublet doublet
2 6.29 doublet doublet
3 6.20 singlet singlet
4 6.05 singlet singlet
5 5.95 doublet doublet
6 5.85 multiplet multiplet
7 5.67 singlet singlet
A second at tempt at isolating B3 in its free form was undertaken this time
on a 200 mg input scale. Once again no pure sample was obtained
al though again there was evidence o f the presence o f procyanidin B3 when
analysed by 400mHz NMR. Another at tempt on a even smaller scale (100
mgs) was undertaken. In order to ensure that the retention t imes for the
compounds on both the analytical and seini-preparative columns were
identical a new 30cm x 3.9 mm id pBondapak C18 analytical column was
used and the following assay was employed for both analytical and
semi-preparative work:
Column: (.iBondapak C18 30cm x 3.9mmid; 10p.m.
Mobile Phase: 90:10 H 2 0 : CH3CN (ApH = 3.0)
Flow Rate: l .Oml/min.
Detection: X = 220nm.
106
Reten t ion Time: Time (Mins.)
Unknown#! 12.48
Unknown#2 15.92
Unknown#3 18.11
B3 20.27
Unknown#4 23.28
(+)-Catechin 25.15
Unknown#5 34.62
Unknown#6 38.27
Unknown#7 44.19
Unknown#8 48.30
From analysis o f the freeze-dr ied product it was evident that this reaction
produced many products and that as for the acetylated reaction it was
apparent that the yield was much less than the stated 50%. The reaction
profile for this p roduct was summarised as follows:
Table #29: Reaction Profile for Free Hydroxy (Fonknechten Reaction).
Sample/Ret
Time
ukl
12.48
uk2
15.92
uk3
18.11
B3
20.27
uk4
23.28
Cat 'n
25.15
uk5
34.62
uk6
38.27
uk7
44.19
uk8
48.30
Product 0.14 0.41 0.19 21.99 20.2 32.3 11.03 2.45 3.26 7.03
Standard 0.19 n/d n/d 96.15 n/d 0.21 n/d n/d n/d n/d
The scale up to semi-preparative HPLC gave the same retention times for
the peaks o f interest , thus a direct comparison was possible. The collected
fractions were analysed and those collected over the first 25 fractions gave
the best results with the first 15 being combined to give a peak with a % HI
purity o f 97.80%. Analysis o f the waste aqueous fraction, after ethyl
107
acetate extraction showed that a significant quanti ty o f procyanidin B3 was
not extracted.
The isolated product was analysed by 400mHz N M R and again an impure
spectrum was obtained. Analysis o f the crude product which had
beenretained in solution at ambient for several days showed the profile to
be unchanged therefore suggesting that the compound was not degraded
when in contact with mobile phase. This fur ther reinforced the suggestion
that the compound may have degraded during the ethyl acetate extraction
procedure .A comparison o f the ]H N M R for the authentic B3 and the
collected fraction is summarised in the following table.
Table #30: Comparison o f Isolated Fraction and procyanidin B3 Standard.
Pro ton No: Chemical Shift
(ppms)
B3Std : Suspect
Peak Type
B3Std : Suspect
1 7.35 : 7.35 singlet singlet
2 7.20-7 .10 : 7.20-7.10 doub,doub:doub,doub
3 6.90 6.90 doublet doublet
4 6.87 : 6.87 doublet doublet
5 6 .85-6.70 : 6.85-6.70 multiplet multiplet
6 6.50 : 6.50 doublet multiplet
7 6.46 6.46 doublet doublet
8 6.29 6.29 doublet doublet
9 6.27 : 6.27 doublet doublet
10 6.20 6.20 doublet doublet
11 6.05 : 6.05 singlet singlet
12 6.00 6.00 singlet singlet*
13 5.91 : 5.91 ----- : doublet
14 5.89-5.80: 5.85-5.80 multiplet multiplet
108
* Impurity at 6 .00ppms due to ethyl ace ta te interference
The more r igorous method proposed by Delcour e t a l was then
investigated as it was felt that this synthesis should lead to fewer
impurities, making isolation o f pure product easier. This reaction was
carried out on a small scale (lOOmgs) and from in process HPLC data only
2 peaks were observed. (App.l Chrom#9), one for procyanidin B3 and the
other for catechin. The procyanidin B3 was present at 16.40%II and this
was very close to the expected 20.00% yield. The isolated product had the
following profile:
Table # 3 1 : Isolated Crude Product (Delcour et a l).
Sample/Ret
Time
ukl
8.46
uk2
10.05
EtoAc
11.82
uk3
13.28
uk4
17.23
B3
19.90
uk5
22.0
Cat 'n
24.03
Crude Product 0.02 0.06 2.48 0.40 0.10 16.47 1.03 79.37
None o f the later eluting peaks observed in the Fonknechten et a l method
were present in this profile and this served to confirm that the more
controlled reaction would give a cleaner profile (App. A Chrom#10). The
product was passed down the semi-preparative column but analysis o f the
collected fractions showed that none were o f sufficient purity to warrant
further isolation. The remaining portion o f this crudc Whs purified using a
two step isolation, the first simply to remove the excessive (+)-catechin,
the second to isolate the procyanidin B3. The collected material had a
purity o f 85.00%H1. This was compared to the authentic procyanidin B3 as
follows (App.B NMR#14):
109
Table # 32: Comparison o f procyanidin 133 Std and Product .
Pro ton No: Chemical Shift
(ppms)
B3Std : Product
Peak
B3 Std
Type
: Product
1 7.35 : ----- singlet ------------
2 7.20-7.10 7.05-6.95 doub,doub:doub,doub
3 6.90 : 6.90 doublet doublet
4 6.87 6.87 doublet doublet
5 6.85-6 .70 : 6.85-6.70 multiplet multiplet
6 6.50 : 6.50 doublet multiplet
7 6.46 : 6.46 doublet doublet
8 6.29 6.29 doublet doublet
9 6.27 6.27 doublet doublet
10 6.20 6.20 double t doiblet
11 6.05 : 6.05 singlet singlet
12 ------------ 6.00 ----- : multiplet*
13 5.91 5.91 — doublet
14 5.89-5.80 5.85-5.80 multiplet multiplet
15 5.65 : 5.65 singlet singlet
* Peak due to Ethyl Aceta te Interference
110
This was the last o f the a ttempted syntheses o f procyanidin B3 in its free
form by these synthetic methods. In all cases these reactions probably
suffered from small scale input o f (+)-taxifolin due to the expense o f this
material. The poor return for procyanidin B3 makes either o f these
procedures examined impractical. However , it should be noted that scale
up o f both procedures may have an advantageous effect on the synthesis.
As a result o f this work it became clear that a more suitable form o f
synthesis would be one that employed an enzyme to effect dimerisation o f
catechin. This is a common react ion in nature where catechin is oxidised
to its higher molecular weight oligomers by the action o f several oxidase
enzymes. Based upon l i terature it was decided to attempt a bio-mimetic
reaction to form procyanidin B3. Much work has been done into the
identification o f the mode o f action o f these enzymes and it has been
suggested that one o f the peaks present in these reactions was procyanidin
B3 [4],
Procyanidin B3 by enzymatic oxidation:
The initial procedure that was examined for the enzymatic oxidation of
(+)-catechin was that described by Oszmianski et al. This group reacted
chlorogenic acid, catechin and a mixture o f both with tyrosinase, a
polyphenol oxidase enzyme extracted from mushrooms. The react ion was
allowed to proceed for a 12 hour period and was s a m p W at T = 0, T= lhr. ,
T = 2hrs. and T = 12 hrs. The reaction was monito by HPLC at both
220 and 280nm and a compar ison between these w;* lengths was made.
The pH o f the acetic acid solution in which this ¡c;iction was carried out
was adjusted to 3.50 as it was felt a more controlled reaction would take
place at lower pH's. The reaction profile for this experiment was
summarised as follows:
111
Table #33: Enzymatic Oxidation Reaction Profile A, = 220nm.
Sample/Ret
Time
ukl
6.24
uk2
10.17
uk3
11.40
uk4
12.85
uk5
14.70
Cat'n
20.32
uk6
28.48
uk7
29.28
TO post Add'n 0.22 n/d 0.04 n/d n/d 98.60 0.96 0.15
T = lH r hold n/d 1.60 n/d n/d n/d 98.40 n/d n/d
T = 2Hr hold n/d n/d 1.86 0.81 n/d 95.92 n/d n/d
T = 12Hr hold n/d n/d 2.65 1.30 0.51 95.54 n/d n/d
Table #34: Enzymatic Oxidation React ion Profile X = 280nm.
Sample/ Ret
Time (mins).
ukl
5.54
uk2
10.62
uk3
18.11
Cat'n
18.72
uk4
20.70
TO post Add'n n/d n/d n/d 100.0 n/d
T = lH r hold 3.73 n/d n/d 96.27 n/d
T = 2Hr hold 12.32 n/d n/d 85.47 n/d
T = 3 Hr hold 6.95 n/d n/d 93.05 n/d
T = 12Hr hold n/d n/d n/d 100.0 n/d
B3/Cat 'n Spike 1.41 15.07 1.90 81.11 0.50
From this initial experiment it was clear that very little react ion had taken
place and that procyanidin B3 might not have been present. Since the
experiment produced very little react ion products at pH 3.5 it was decided
to repeat the procedure at the higher pH o f 6.5. It was also observed that
the wavelength o f 220nm gave a better response for all products monitored
and this was now the wavelength o f choice.
112
Table #35: R eac t io n Prof i le fo r O x ida t ion @ pH 6.5.
Sample/Ret
Time(mins).
ukl
5.10
uk2
5.80
uk3
7.50
uk4
8.20
uk5
9.50
uk6
10.28
uk7
11.20
uk8
12.70
uk9
13.80
Cat'n
16.30
TO post Addn n/d n/d n/d n/d 0.04 0.12 n/d 0.1 0.02 97.32
T = 5mins n/d n/d n/d n/d 6.41 1.94 n/d 0.3 0.55 90.2
T= lH r hold n/d n/d n/d n/d 7.26 1.66 0.29 0.6 1.35 86.9
B3/Cat spike 0.06 0.93 0.22 0.38 22.2 1.12 6.4 1.3 0.8 66.6
T = 2Hr hold n/d n/d n/d n/d 7.2 1.8 0.25 n/d 0.75 89.6
T = 0 / N hold n/d n/d n/d 0.4 6.7 2 n/d n/d 0.4 88.9
The pH adjustment to a value close to neutral did appear to have a positive
effect on the reaction. A peak with the same retention time as procyanidin
B3 was observed after just 5mins. contact with the enzyme. However, it
was also observed that this peak did not appear to g row significantly with
time. Yet another experiment used the same parameters with the exception
o f the enzyme concentra tion which was doubled. The react ion was
observed to form a dark yellow colour almost immediately which turned to
brown over time. The react ion profile for this experiment was summarised
as follows:
Table #36: Reaction Profile for Increased Enzyme Concentration.
Sample/ Ret
Time(mins).
uk l
5.27
uk2
9.57
uk3
12.54
uk4
13.26
uk5
14.56
Cat 'n
17.14
TO post Add'n 0.32 0.34 0.14 n/d n/d 98.40
T = 5mins hold 0.37 8.76 2.46 n/d 1.47 86.93
T= 2Hr hold 1.56 6.56 0.95 0.90 167 88.35
113
The results obtained suggested that a l though colouration did appear to be
more rapid for higher concentra tions o f enzymes this did not mean a
corresponding increase in the desired product peak was obtained. A
comparison for the two 5min. traces showed a similar react ion profile in
both cases. (App.A Chrom 11).
Another feature o f these experiments was the fact that a slight decrease in
the % area for the unknown product was observed over time. This may
have been due to the fact that the enzyme was further reacting with this
formed product to form higher oligomers. Following on from these
experiments it was decided to scale up the react ion and to attempt a
separation o f the unknown peak by semi-preparative HPLC. For the
reaction 600 mgs. o f (+)-catechin was used. The peak o f interest was
observed to have a %II o f 12% in the isolated crude material following its
extract ion from ethyl acetate. Following the first purification at tempt it
was observed that a large concentra tion o f catechin still remained and that
this material warranted a fur ther purification. The reaction profile after
the first purification was summarised as follows:
Table #37: Reaction Profile o f Purified Enzymatic Oxid'n Product .
Sample/
RetTime
u k l
5.36
uk2
5.62
uk3
6.33
u k4
7.18
uk5
8.48
uk6
10.2
uk7
11.3
uk8
13.1
uk9
14.0
uklO
14.8
Cat 'n
17.9
u k l 1
18.7
pdt 1.17 4.81 1.73 0.2 1.05 28.8 2.49 0.44 0.49 7.46 44.7 6.31
Even though the react ion appeared to proceed to yield a product which
corresponded to procyanidin B3 retention time, there was insufficient
concentra tions o f it to allow proper isolation by HPLC. This reaction was
repeated with the same results, except that the isolated product following
the two extractions had a % HI figure o f 94% by HPLC. The material was
immediately analysed by 400mHz N M R and the 'H spectrum showed that
the product was not as pure as first thought. However, from this spectrum
it was not possible to identify any o f the key chemical shift resonances
114
attr ibutable to procyanidin B3. This sample was evaporated to dryness
and reconst i tuted with mobile phase and analysed by HPLC in order to try
and identify some o f these impurities. It was clear that about 5%II of
these impurit ies was due to the presence o f catechin and since this was not
apparent in the frac tionated samples it was most probably due to
degradation. (App.A Chrom #12).
Since this reaction appeared only to produce the peak o f interest in 8 to
10% yield another approach was needed in order to identify this peak. It
was decided to allow the reaction to proceed in an NMR tube thereby
enabling constant monitoring. On this occasion the reaction was also
moni tored by HPLC by removing small al iquots from the N M R tube.
Running the reaction on this scale highlighted one o f the problems that had
already been observed in some o f the experiments. Under these conditions
catechin did not fully go into solution pr ior to the addition o f the enzyme
thereby somewhat reducing the quantity o f starting material available to be
acted upon by the enzyme. To allievate this problem approximately
10%v/v ace tone was used to fully dissolve the catechin before the
deutera ted water was added. The in process HPLC analysis for this
reaction was as follows:
Table #38: Reaction Profile o f Acetone Soluabilised Compounds.
Sample/Ret
Time(mins).
Act'n
5.15
u k l
6.06
uk2
6.64
uk3
7.58
uk4
7.98
uk5
8.80
uk6
11.15
uk7
12.14
uk8
12.88
Cat'n
13.63
T = 20mins 0.7 0.06 0.05 0.16 0.19 26.24 2.42 0.48 1.36 67.9
T = 40mins 0.64 0.21 0.24 0.1 0.05 37.6 2.34 0.36 1.5 55.26
T = 60mins 0.61 0.19 0.32 0.13 0.1 29.3 4.5 1.3 2.79 58.45
115
Il would appear that a product had formed with a retention time o f 8.8mins.
but it was not certain whether this peak was solvent related or a legimate
product. The N M R spectra that accompanied these traces did not show
any sign o f resonances associated with procyanidin B3. (App.B NMR#15) .
The sample was subsequently held at ambient for a considerable period o f
time and was analysed after 20 hrs. and 6 days respectively but failed to
show any fur ther reaction. The react ion was repeated without using
acetone to dissolve the catechin in order to verify that the use o f acetone
was not a problem. The same reaction profiles and NMR spectra were
generated for this experiment. (App.B NMR# 16). Finally, even though
there was no evidence for the production o f procyanidin B3 the fact still
remained that a product was being produced and an effort was made to try
and identify what composi t ion this unknown possessed.
Following on from the experience gained in the isolation o f procyanidin B3
as its acetylated derivative it was decided to utilise the same methodology.
The react ion profile for the isolated compounds following acetylation was
as follows:(App.A Chrom #13)
Table #39: Reaction Profile for Acetylated Enzyme Products.
Sample/Ret
Time(mins)
Cat 'n
12.99
uk l
17.66
uk2
18.08
uk3
18.67
uk4
21.04
uk5
21.92
uk6
25.52
uk7
29.57
uk8
30.74
uk9
33.10
Acety'd Pdt 91.66 0.07 0.15 0.49 0.26 0.06 0.94 1.45 4.7 0.22
The retention times observed for the peaks o f the enzyme reaction agreed
closely to those observed following the acetylation o f the condensation
products formed from the method o f Fonknechten el al. The product was
subjected to purification by semi-preparative HPLC. The collected
fractions were analysed but no evidence as to the purity o f the collected
peak was obtained due to the low dilution at which the compound was
collected. Analysis by 400mHz N M R showed firstly a lack o f product and,
116
secondly, what product was present appeared impure. However, some of
the resonances may have been at tr ibuted to acetylated B3. (App.2
N M R #17).
It was felt that if any product was to be isolated from the enzymatic
reaction then a higher propor tion o f the compound o f interest would have
to be formed. Fur ther experiments were undertaken with the aim o f
increasing the % o f the product formed Three experiments that varied the
enzyme concent ra tion were attempted. The first was a control using the
concentra tion as described, the second used four times the concentra tion
while the third used ten times the concentration. All three were started at
the same time and allowed to stand at ambient for 2 hours with constant
agitation. They were all quenched in the usual manner and acetylated.
HPLC analyses for the three isolated materials was summarised as follows:
Table #40: Reaction Profiles o f Alternative Enzyme Cone. Products.
Sample/ Ret
Time(mins).
syst'm
1.82
Cat 'n
11.20
uk l
16.48
uk2
22.68
uk3
25.68
uk4
26.60
uk5
28.76
uk6
37.76
uk7
47.82
uk8
60.03
Control n/a 65.9 0.8 10.89 1.09 7.89 1.44 0.6 1.67 4.11
4 X enzyme n/a 28.55 1.16 17.34 n/d 6.24 3.15 6.3 23.63 9.49
10X enzyme n/a 34.85 3.01 9.59 n/d 8.53 5.27 7.82 17.57 n/d
It was apparent from these results that the addit ion o f extra enzyme did not
lead to an increase in the peak o f interest but rather an increase in later
eluting peaks assumed to be higher oligomers. In all cases the %II area of
the peak was similar and suggested that a finite level for this peak had been
reached. It was also noted that a higher concentra tions o f the enzyme
117
caused a rapid precipitation o f the products upon quenching thus limiting
the quanti ty o f material recovered. The material from the control
experiment was analysed by 400mHz N M R and yet again no evidence o f the
appearance o f procyanidin B3 was found.
Two more a ttempts at isolating this unknown product were undertaken.
The first o f these was a repeat o f the previous reactions in every detail with
the exception o f the extraction o f the crude product into ethyl acetate. In
this case the aqueous mother l iquor was removed by ro ta ry evaporat ion
directly in an effort to limit product degradation. Analyses o f this mother
l iquor with 50% o f its volume removed showed that the product was
present at 61%HI, with catechin at 10%II and several other peaks making
up the remainder. Instead o f subjecting this recovered material to a
second purification it was analysed directly by 400mHz NMR. There was
no evidence to suggest that the peak o f interest was procyanidin B3
although some o f the resonances common to catechin were found as would
be expected. (App.B NMR#18). The second experiment did employ a
second purification, however, as had already been seen the concentra tion of
the collected fractions was too dilute to allow analyses. The collected
product was o f a good purity and the 'H N M R failed to show any o f the
resonances associated with B3. The profile o f this products was as
follows:
118
Table #41: 'H N M R Profi le for Iso la ted E nzym e Produc t .
Peak Type Chemical Shift
(ppms)
Signal Type
1 X H-Ar 8.60 doublet
H-Ar 7.80-6.65 multiplet
----- 6.60 singlet
----- 6.45 singlet
----- 5.90 doublet
----- 5.50 doub,doub
----- 5.20-5.05 multiplet
---- 4.20-4 .00 multiplet
Some o f the shifts observed were also seen in traces for procyanidin B3 but
the characterist ic doublets at 4.78 and 4 .30ppm were not present further
reinforcing the thought that procyanidin B3 was not present. Finally, two
alternative experiments were carried out whereby the starting material was
dissolved in 10%v/v CH3CN prior to enzyme addition. The reason for
using this solvent was to maintain all the compounds o f interest in solution.
The first o f these experiments used a mild acid wash to remove pyridine
post actylation while the second was a repeat but omitted this wash jus t to
confirm that the acid had not caused degradation. No evidence for the
presence o f procyanidin B3 was obtained in either o f these experiments.
Ancillary to these experiments react ion blanks which omitted (+)-catechin
and the enzyme respectively were undertaken.
119
These two experiments served to prove that the mere presence o f either
catechin or enzyme was insufficient to produce the observed peak. All o f
the experiments completed, seemed to produce a reaction product with a
retention time similar to procyanidin B3 in 8 to 10% yield. Isolation of
this material proved difficult but NMR evidence served to discount the
suggestion that this peak was in fact procyanidin B3. Following a review
o f the l i tera ture a reference which suggested to the identity o f this
compound was obtained. In the 1960's Weignes et a l exposed (+)-catechin
to peroxidase enzymes and proposed that the resulting product was in fact
dehydrodicatechin A, a (2P-0 - 7 ) linked dimer o f (+)-catechin. The
supposit ion that this was the unknown product made some sense as this
dimer would probably have similiar retention time on HPLC as procyanidin
B3 but would certainly have a different N M R profile.
The first at tempt at this reaction was undertaken in duplicate and was done
so to compare the enzymes tyrosinase and horseradish peroxidase. Both
reactions failed to produce the yellow precipitate as described after
standing for 14 days. Re-examination o f the li terature indicated that
during the 14 day hold it was necessary to add fresh enzyme and hydrogen
peroxide on a daily basis. The following experiment recreated the method
as laid down and after 14 days a yellow precipitate did form. HPLC
analysis o f the reaction after 7 days showed primarily unreacted catechin
but also a peak with a similar retention time to that o f the unknown product
and procyanidin B3. (App.A Chrom#14). Having followed the extraction
procedure the isolated product was analysed by HPLC and was found to
mainly contain (+)-catechin. (App.A Chrom#15). Analysis o f the waste
mother l iquor showed that what product had formed had failed to extract
into the organic layer thereby explaining why so little o f this material was
isolated (App. A Chrom#16).
120
Some question as to the activity o f the enzyme was raised as this material
had been stored at -20"C for a considerable period o f time. The in
process HPLC details for this react ion were summarised as follows:
Table #42: Inprocess Reaction Profile for Enzyme Mediated Products .
Sample/ Ret
Time(mins).
uk l
2.59
uk2
3.08
uk3
6.40
uk4
10.78
uk5
12.91
uk6
17.15
uk7
18.86
uk8
20.59
Cat 'n
23.18
uk9
28.08
T = 7 days n/d n/d n/d n/d 0.17 0.18 6.89 1.72 89.61 1,41
Isolated Pdt 0.12 1.77 0.4 0.28 n/d n/d 1.54 0.43 95.17 n/d
Waste M/L 0.16 n/d n/d 0.21 n/d n/d 10.77 2.18 82.74 2.52
The enzymatic format ion o f dehydrodicatechin A did appear to produce a
peak with the same retention t ime as procyanidin B3. Although this
material was not isolated and identified it is possible that this was the
product produced by the o ther enzyme reactions.
121
CONCLUSION:
From all o f the experiments carried out in the examination o f the synthesis
o f procyanidin B3,a few interesting observations were made regarding its
isolation. Firstly as had been stated in the l i terature the synthesis o f the
condensed tannin dimers and tr imers was not a trivial procedure, in fact
very few groups had ventured into this area o f chemistry, as it was frought
with difficulties. In its free form the isolation o f procyanidin B3 was not
achieved using the appartus and conditions investigated. This does not
mean that it was not possible to isolate it but rather that it was a labile
material that required special treatment during extraction. Secondly the
analytical conditions that were developed worked well for separating the
compounds and even aided isolation o f the acetylated derivatives, but they
may be fur ther enhanced with the use o f gradient elution facilities. Also it
was " ' ' served that the percentage yield obtained in these syntheses did not
justify the type o f cost in equipment and solvent that would be involved in
the recovery o f the dimer. The use o f 400mFIz NMR was o f great
assistance for the identification o f the compounds made and also gave a
good indication o f the purity o f any isolated compounds. Finally we were
unable to confirm li terature reports o f the presence o f procyanidin B3 in
the enzymatic oxidation o f catechin.
I f any further synthetic work was to be carried out 011 these compounds a
ready supply o f the start ing compound (+)-taxifolin would be necessary.
To this end it may be benificial to try and synthesise this compound
initially, with the possible aim o f telescoping the procedure and purifying
the resultant products by commercial scale preparat ive chromatography. I f
the procyanidin B3 could be improved it would be o f interest to try and
identify the decomposit ion products o f the dimer in alcoholic solutions
possibly by LC-MS, as this would go a long way towards unders tanding the
modes o f action involved in chill haze formation in beers and the ageing o f
wine.
122
REFERENCES:
[1] Delcour J A., Ferreira D , Roux D.G., ./. Chew, Soc. Perkin Trans. J,
(1983) , 1711.
[2] Fonknecliten G., Moll M., Cagnianl D., Kirsch G., Muller J .F.,
./. inst. B rew ., 89 ,(1983), 424.
[3] Delcour J A., Vercruysse S.A.R., J .ln s i. B rew ., 92, (1986), 244.
[4] Oszmianski J., Lee C.Y., J. Agric. Food. C hew ., 38, (1990), 1202.
[5] Weignes K., Eberl W., Hulhwelker D., Mattauch H., Perner J.,
L ieb igs Ann. Chen?., 726, (1969), I 14.
[6] Outtrup H., Schaumburg K., C arlsberg Res. Coinm im ., 46, (1981) , 43.
123
A P P E N D IX #A: 1IPLC C H R O M A T O G R A M S
124
pu i
5 .28
CATMEC1N 1,4.1
FILE 1
PEAK*
123
METHOD 0.
AREA*
2.22 9 87.02 10.751
RT
12/01/91 11:54:03 CH= "A* PS= J.
RUN 24 INDEX 24 BIN 21
AREA BC
3.28 119477 026.81 4664062 02!1.1 576241 03
TOTAL 100 . 5359780
App. A, Chrom #1: Isolated unknown post semi-preparative
pur if icat ion.
r
App. A, Chrom #1: Unknown spiked with catechin.
er e 20DATA SAVED TO BIN * 22
CATHECIN 1,4. 1 12/01/91
FILE 1 . METHOD 0. RUN 25
PEAK* AREAZ RT AREA BC
1 0.091 3.31 30154 022 0.179 3 .9 59685 023 13.871 6.8 4613772 024 85 .859 10.99 28558585 03
TOTAL 100. 33262196
INDEX 25
CH= 'A* PS= 1 .
BIN 22
X I 0 1 m inu tes
it
App. A, Chrom #3: Guinness Group Research analysis o f unknown.
3
10 .00
> 14 .70
App. A, Chrom #4: Purified fraction o f procyanidin B3 suspect.
DATA SAUED TO BIN « 23
CATHEC1N 1,7
FILE 1 . METHOD 0 . RUN 26
PEAK# AREA* RT AREA BC
1 0 .019 3 .65 2690 012 0 .244 6 .72 33700 023 95 .648 8 .48 13203595 024 3 .179 10. 438813 035 0 .909 14 .7 125513 01
0 2 /2 5 /9 2 17:30 :35
INDEX 26
CH= 'A ' PS= 1 .
BIN
TOTAL 100 . 13804311
4
II00CL.
IIXo
r-ro
MOn\r-CM\(O
ztn
xUJQZ
z3£X
QOXh*UJ
z
oUJXI-<ru
uact
<JLUJoc<x
XaUJoc<1
»<xUJQ .
CM (M CM CM CM ro CM CM CM ro® ® ® ® ® ® ® ® ® ® ® ® ®
00 ® eg r - O« **■ r - O 'CD CO CM in tn >0 ro <o N O ' m *— ©
ro ® «—1 o o» *■ m 'O 00 ro m O '•O fO ro •-• O ' O ' m CM «o CM fO O ' vO
•— CM ■< ■ r w— CM m m CM vO00 ro O '
fO
m CM ro * © >0 CM © o■*■ 'O o ro ro o . in *n « r •< CM
uo © O ' — CM ro ‘O © O '
<0 CM O ' ’O sO CM ■'Tro *0 r - * ro ro O ' £ O ' m © ■<
® to in 'O ro ro m 00 ® r -
® a ® ® *o ® ® to ® ® ®■< •“ ro ®
<1f-o
App. A, Chrom #5: Isolated product post 1" analysis & after
ethyl acetate extraction.
Hi/iO ,
IIXCJ
mCM
CMm
MUJ
® QZ
«—« —«—
CM l> — C M C M C M C M N C M C M ©O' © ® ® ® 9 ® ® ® ® ®\s m <1 < c ® r ' * f o m r o * - m ^CM CM UJ r o ' o o v ' - r o m a o - o r -\ oc C M O ' ^ ' O O ' — — r o ®ro <1 •— ' 0 © 0 ' 0 ' © © © » -® Z » - ' T *0 "O ® *0
Z> ro ® O ' ^OC CM —
aCM ®
QO
Z Xt—
© UJ Xr <1
o UJH r - oc
<1O •—UJ3 z<r00 a
UJ •*<r X UJ X .j - k- — j <1
<i uo o u . Ou
ro(Sicn
'■c so © >o *-«ruor-r^©® —
00 *,O'CM'O©*»n0' ® .— m(o®inc0fo® ® — ® ro r* — ^ r*o®®©csir^sON®ao
5,~.C'4rT)‘*ri/*>'OP-000N <rh-O
App. A, Chrom #6: A ce ty la ted p ro d u c t s ex F onknech ten react ion.
r s j r o r o r o r o — — — — — — — — — —co ro — ® ' O o o ^ j O ' u » > . t u r s j — ® * o o o ^ j f f « c " . * c o r o —
D ® ® ® ® ® — ® ® ® ® ®N* ® ~ ® 9 9 6 ® ® ® ® ® ®
D00
MJk j -
WC"r-j
fvj*C
.16
8co®'O
COC"
‘■Sjro00
OP
s— vl 'O 0*
UlC* '9. « ®
03 N 00 S> C7« PO
— ® 00 ® ® ®Oo -0 — V* tu OP CM ' O 00 'O
- ® ®co fsj ® < 00
i>
2>»mX
0 ro hi N r » __ m» mmIf c* ro ® >0 «sj (7* 0 CO N ro — ® 00 05 ^ 0s O 0s 1/ L" y i «Fts L . ^0 O' ® L . ® bo OO ’■sj ^ 00 > %i !* ® a« ■si ^ rsj ® 1» —
aJH
“ *00 >■ ’ — ro OJ O' * fr* 'O N 00 cn N) c* > ro
ro - _ ■ AV" — CO -
cu ro O' OJ j- c" cm A CO 00 M Ol — »— — <7» —■ ro% ® L* £► •—» o j 00 C" S3 c* ro *0 00 00 1* 00 ^ N to ® -si O" *■ — »5 ' J -sj Uf 'C ro >> ® OO ® —» ® CO N U" U» Co VI CO 00 CO X »N 00 O) •— -si -*>1 - j a OJ c* CO ro <7- >0 — CO W Co 0 -g ® ® 00 C" ni5 — — <7* ® co 'O OD .Jfc -K ■X CM ® fSJ 00 <7* ® CO OJ -vj ® -O 'O -sJ 3>J ® ® ® ® ® ® ® ® (S ® fi ® ® ® ® ® ® ® ® ® ® ® ® CO) ro OJ ro ro ro r\3 ro ro rsj ro N N fSJ — <0 ro ro o j ro n 0
r*m
3m-t3:oo
30CZ
ZOmX
oI>
3>OmH<rx>
fO
ro
3
oXII
s- I1>V)
ImoHO00ztf
iT■-o®
03z
6
CHANNEL A INJECT 0 3 /0 8 /9 3 12:02:54 STORED TO BIN I 3 .04
DATA SAVED TO BIN # 3 App. A, Chrom #7: Authentic procyanidin B3 peak marker.
(Guinness Group Research)
FREE PROCYANIDIN B3
FILE 1 . METHOD 0 . RUN 3
PEAK# AREA* RT AREA BC
1 0 .023 0 .04 1334 012 0 .873 2 .4 51517 023 1 .199 2 .83 70702 034 0 .071 3 .76 4183 025 0 .421 4.14 24861 036 0 .236 4 .84 13946 017 1 .764 5 .72 104048 018 0 .33 7 .68 19462 029 94 .541 9.1 5576494 02
10 0 .542 10.86 31961 03
0 3 /0 8 /9 3 12:02:54
INDEX
CH= 'A ' PS= 1.
BIN 3
TOTAL 100 5898508
1 -n -ns t-n — »* 1> rk *-» —» :* m m
ro 0 'D •j u\ CO ro at00
— ijj3>CO• :[> CO
> 'O> CD c© <9 l© ® —» <o <o tO CO <o m •<
i> 3co (O —* -* to tO ro —* CO n m«O 1© cn -o •o © UJ Ht/i 0 ID (0 ^ UJ ^ O' *** -K J
O
to CO Co cO® (-n CD -vl r*> 10 CO a> *sj c-n X
:o<E <» (0 -vi ~nJ r o -fr. CO CO COCT' -HCO ro ro UJ t/1 UJ — <7t ‘ J O 'O*1J CsJ -Nl tJJ ro *© CO cJ CO '© *—CO C\J to (» Jh. -J 'O 2* c* CO si ro '0 'O <£> c/1 CO Jfc 1I> ~ a> no '0 ® '0 'O <D t> on CO \1 o ® *—9 Ul O' ® CD m CO1 f-0 ro •si •—— —• ® CD CD CD\<c ® to <o > 0 (0 (0 to CO CO cO cO to ■o
*— Ul 03 CO r o oj cm CV3 tO CO o to
XOmx■si
oXIICD *'0— COZ II
rnoono
z
HO
.\
® 00® CD — \ 0 © C O - - -
^ ■‘O cu ro ® ® co— o ^ ® ® si cn >o
OJ o >0 o co — ro
so o co u) co ro ro ro
— a> Ln Kj ® 'O si 0Koo ro oo cn o> o
lllr __O' *cn O ' u> ■K — NCD «> ro co cn to “si -o coro 9 on ® ®cn O ' t n Jk rs> O ' 00 CO CDO' ® 'O a* — -vj O" ro cn
© ® ® ® ® ®co w ro ro n n
k ^i i
■T3nxM
I>-jom 2mHXoo
wss
30m
CDCO
>COCOT><
App. A, Chrom #8: Puri f ied f rac t ion o f procyanid in B3 in free form,
co m p ared to au then tic peak marker .
OXII
CD
z"OGOII
8
DELCOUR FREE B3
FILE 1 . METHOD 0 . RUN 1
PEAK# AREA* RT AREA BC
1 0 .055 2 .6 10428 012 0 .059 3.03 11070;[913 0 .289 3 .8 54406 014 0 .06 9 .36 11393 015 0 .255 13.34 48096 016 16 ,36 23 .15 3080911 014-7 82 .921 28 .59 15615708 01
TOTAL
01/08/94 12:50:58
INDEX 1
CH= ’ A" PS= 1.
BIN 1
11
$1 * Jb "
100 18832012
/TUrKft n. tire * o : lO-iw^LÇ
^ J , c j A
DELCOUR FREE B3
FILE 1 . METHOD 0 . RUN 4
PEAK# AREA* RT AREA BC
1 0 .033 2 .59 21042 012 0 .0 0 5 3 .04 3145 013 0 .005 3 .57 3147 014 0 .034 6 .96 21721 015 0 .021 8 .46 13643 016 0 .058 10 .05 36776 01
01/10 /94 17 :20 :58
INDEX
CH= ’ A' PS= 1 .
BIN 4
7 2 .48 11 .82 1583865 028 0 .404 13 .28 258199 039 0 .097 17 .23 61897 02
0 2 * - P310 16 .464 19 .9 1051600111 1 .034 22 . 660443 02 912 79 .366 24 .03 50694678 03
T O T A L 100 . 63874557
INPUT OUERRANGE AT RT 2 .58
ENZYME CAT 1 ,3 05/15 /92
FILE 1 . METHOD 0 . RUN 3
PEAK« AREAX RT AREA BC
I 6 .414 8.97 275885 022 1 .943 9 .48 83580 033 0 .2 9 9 12.34 12870 014 0.311 13.52 13362 025 0 .241 14.24 10384 026 0 .536 15.02 23034 027 90 .181 15.87 3878704 088 0 .075 17.63 3207 05
CH= 'A'
INDEX
PS= 1 .
BIN 3
TOTAL 4301026
10
LrilHllM Ll- n «>M 1 IZ 1
ECAT 1,6 PREP2 06 /13 /92 16 :20 :27 CH= *A' PS= 1.
FILE 1 . METHOD 0 . RUN 30
PEAK# AREA/. RT AREA BC
1 0 .142 4 .87 58165 012 0 .147 5 .88 60378 023 0 .723 6 .54 297077 024 84 . 7 .6 34497279 025 5 .345 9 .18 2194915 026 3 .879 9 .66 1593225 03
7 5 .133 12 .41 2107912 028 0 .447 13 .64 183462 029 0 .184 14 .28 75582 03
TOTAL 100 . 41067995
12
.08
^ ( i ^ ¿ S o - * *
^ J U U | J L * v s o \ s o A c » '* te > ( f y H > 0 )
% U 7 t U t • >'D ^ | ^
*■ A ' - o*cXOv,
12 .99
.04App. A, C h ro m #13: A ce ty la ted enzym e reac t ion products .
5 .52
fo-8>4<- ? 30 .74
13
10/21/92 16:43:46 CH= "A* PS= 1.
r>i.... ^ i MnrY 7 B1N 1
1
5 .60
C h ro m #14: D ehy d ro d ica tech in A fo rm at ion f rom H o r s e Radish
Peroxidase .
C jttLko'VVA'K ^-----------------------------------------------------Z37T8
DEHYDRODICATECHIN A ASSAY 02/04/94 16 :59 :2? CH=*A* PS= I .
FILE 1. METHOD 0 . RUN 2
PEAK# AREA* RT AREA BC
1 0 .0 1 8 5 .6 3377 012 0 .1 6 6 12.91 31148 013 0 .182 17.15 34143 024 6 .892 18 .86 1294396 025 1 .722 20 .5 9 323503 026 89 .613 23 .18 16831447 087 1 .408 28 .06 264382 05
rOTAL 100 . 18782396
App. A,
14
DATA SAVED TO BIN # 1
DEHYDRODICATECHIN A ASSAY 02/18 /94
FILE 1. METHOD 0. RUN 1
PEAK* AREA* RT AREA BC
1 0 .124 2.59 25799 012 0 .022 2 .92 4530 023 1 .775 3.08 369090 084 0 .054 3 .77 11229 065 0 .101 4.04 21081 076 0 .029 5 .48 6004 017 0 .398 6 .4 82827 018 0 .069 8 .69 14408 019 0 .285 10 .78 59238 01
10 1 .546 18 .08 321357 0211 0 .427 19 .87 88709 0212 95 .169 22 .48 19784892 03
TOTAL 100 . 20789164
16:15 :58
INDEX
CH= 'A* PS= 1.
BIN 1
15
App. A, C h ro m #16: W as te m o t h e r l iquor pos t ethyl ace ta te ex t rac t io
18 .38
DEHVDRODICATECHIN A ASSAY 02/18 /94 16 :47 :24 CH= '
FILE 1. METHOD 0. RUN 2 INDEX 2
PEAK* AREA* RT AREA BC
1 0 .156 2 .53 17522 012 0 .008 2 .89 861 013 0 .028 3 .32 3166 024 0 .326 3 .46 36680 025 0 .085 3 .78 9521 026 0 .312 4 .05 35190 037 0 .095 4 .65 10702 018 0 .092 5 .52 10386 019 0 .212 10.78 23839 01
10 0 .466 16.51 52454 0211 10 .774 18.38 1213678 0212 2 .185 20 .24 246118 0213 82 .741 22 .91 9321097 0314 2 .522 27 .52 284125 01
' PS= I ,
BIN 2
TOTAL 100 11265339
A P P E N D IX #B: 400 mHz N M R S P E C T R A .
125
PPM
5 Oi
4-c'! . I
App.
B,
NM
R #1
: 'H
NMR.
of
(+)-
Cat
echi
n (D
20).
l*< yIn 5
»—
-4 a ;v •UJt- 0
^ |M
•w4.i- .4 % 0 .« M O in • rv in a
1 * °t v r
)t! \
\ M 1 o
1/
N 0 2 6 0 F . i 04DATE 27-3 1-91
SF 400 135SV 230 00! 6056 47JSI 16384TD 16384bri 15151 515BZ/PT 1 B50
PH 3 0RD 1 000A3 541RG 40NS _ 16TE " ^ 9 7
FM 1900002 0 0DP 63L P0
LB 1 000SB 0 0CX 35 00CY 0 0FI 7 4 9 4 0F 2 J 76/0
'm 6=;. 474PPM/K'm :64SR 5443 Q5
i . 0 2.5 2.0
PP
H
i T i 1 i 1 i 1 i 1 i 1 i r1 l1". U. 1. ! . . r.'* 10 :•
TT'jr i’i'T’TT
bò< jk< r
AP A U 4 N 0 E .012 DATE 30-11 -Sa 1
SF 1C0.Ó14 SY 100 . 0 01 1607.<3-7ST 6553?;ID 6553b SW 25 000 .COC HZ/PT ,7bi
PW A 0HP 2. 000AO 1 .311RG 200NS 2C0TE C997
FH 31.300'a i 640"1 > 1DP 17H TP D
LB3B 0 Ö:\ 3 ^ C ‘j:y C 0FI 153 ^ 0 <PF2 23 r.43PHZ/ OM A O Í 7o4PPM/CM 3 q&aSfl -=85G L /
_________U______________J \_____ A_
B & t ' K a e R
. c;■ ■ N0261120.'1MX■ a u p r o g .' ?;:7 At'• DA TE 27-11-91*
312 1024SI 1 512SW2 1992.032SW1 996.016
- tNDO 1
HDH2 S. wnwi S
S5B2 0r 5!>B 1 11
MH2 M‘ PL IM now.' I- 1 7 19hl>- in 2 . 229P- ANT :o l u m n .
& . '■ I 1 7 19bPF C 2 :29P
" ni 0700000p i ? 00
r RG4nr 0 0PH 0 0
- LIE 316 30- NS 4. 1
r , t no OLOOO.iOp i 4 . ‘¡ONF r?e. N O O O cil..O
J=to<uCO
CJ■+
o
VDoU■QCN
m=tfc
£IN03
Q.n-<
7
( PK
Udd
hdd
2£
^
09
0-
(./
OU
Où
001
App. B, NMR #4: C-H Correlation o f (+)-Catechin.
•-* Z ~ 0 O Z T t ^ITJTGO'O^O “n -r > “n t» tíz m winm t m ^ Z i.v r
O l-HX
c-o
O •Hj
O •O 3 r »-* u
o o IO c. o x o u* C . J XO M. vD r o>- o oOf0 *J1 n o o • zO CD ■ i— 3 %J- * o o • jj ' o r*71 r j n ji _ _ - .i «-■* ^
i o Ci ( - X X z c /K ': c n c n o > ~n ■ n xc . o - ¡5* O w □ i t H H > V C "0 re T) ►- XI . CD X S- i j l-k i i-k fV H X 3 33 o
M M* I JM
U CD Ik
m 1. 3 D O T ] O “□ O O X C - I U U X J
U i l O X O l ü I>J J l A j O O D u - 1 - C-»-* c_-
i_fc CD Ü1 C 1 O ' • JiI ' J O I v ’ C/î<7> i i v i UD CÌ OÌ i-» f> o o X
~ o X_5 O c r o i k-k
X
œ
H03W
NI
NIMH
JV'.
I bLJ
bbOH
5
H w (O + (Jl O) \JIS © (9 CS (5 (9 (S O
CJI©
©<9
OI©
ro©©
roen©
J I I L J I L
-Oí(O
- s
©©©
toOI©
-p.©©
ACJI©
OI©©
-00
rooo
ro- S iro
(O
-800
OI
©
3©(O
u> s 3 SS© © s i ?i i « Q ® <
s § s?
X
1C 0 0 UIIO
©
0
co ©\J I01 3) A Co o o© Ico© ro ro
Wi-» O
* o m© ID/O OI J~ t-Hen o00 7Z©©©.©©
App. B, Spec #1: Mass spectrum o f acetylaíed (+)-Catechin
ACETYLATED CATHECHIN
-I—I—I—i l I—|—I—r7.8 7.0
T—I—I—I—r 8.0
T—I—[—I I ; 11 I I4.8 4.0PPM
pp B,
NM
R #6
: 'H
NMR
of ac
etyl
ated
(+
)-C
atec
hi
DATE 18-8-82SF 400.134SY 230.001 80S«.473SI 18384TD 16384SN 18181.818KZ/PT 1.850P* 3.0RO 1.000AO .841as 20KS 18TE t2S7F* 1900002 0.0DP 63L POLB .300es 0.0ex 39.00CY 0.0Fi 7.698PF2 1.189PHZ/CX 74.408PPM/CK .188sa 4388.00
Wd
rJ
c / ! ' ^ 3 ; ' ri T i n r i ^ r_ D O T I - 1 Z XJ > T] "0xj d n p ^ m A x œ m " D r b i . n W G D O i
3 : ^\ n
i n i ^ U)CD X -sj CD .b. 1ftt_k ^ O ro u . rvjLu ■>J *-* >J tu o o CD ^ OO ^ K ^ C T I O C I I O O o o o v j ce o ^ r o £*
TJo ' j o i ^ a - o o o o o o >-* o oO i-»- iD -J U3 o o œ o
"vj JD O o o ot ; T3
X LT—I if? o O" Ln o "Dr^ Î . o ►—i i-k -<. ■n >\ —1ï>T) i-» m O—i ro cn cn o O
'vj i_n i_n o k->- >"-i Ul LF.O) k-k CD
Ll o o o 1"-J cn cr- -j o CD o
i oC 'J CD CD CD »-*is. '"sj r j rv f\jcc CP ij.
App. B, N M R #8: C -H C o r r e la t io n o f ace ty la ted (+) -Ca tech in .CH com FOR ACETYLATED CATHECHIN
i
t
I1
4,t
AU182101.SMX FI PROJ:
PROJH1.0 0 1 F2 PROJ:
PR0JX.001 AU PROS;
Z28.AU DATE 18—8—92
. S I2 409B4 H SI1 B12
— 3M2 1 1 6 2 7 .9 0 7SMI 1 2 3 2 .7 4 2NDO 2
- 2 .0
MOK2 69 K MOM1 6
— c . D LB2 2 .0 0 0LB1 2 .0 0 06B2 0 .0
4 A 881 0 .0— 3 .0 MC2 N
PLItC ROW;FI ~ 130 .212FF2 14 .6 4 2 ?— 3 .5 AND COLUMN:FI 7 .4 3 9 f
- F2 1 .277F— 4 .0 01 1 . S18000C- S3 OH” PI 1 1 .8 0_ DO .000003Cimi. 4 .8 PB 1 1 .0 0- D2 .003450C■ PS 5 .5 0
D4 .001720C_ S.O S2 17H- R6A“ RD 0 .0
PW 0 .0_ B.B DE B 6 .3 0
US 64DS 2P9 9 5 .0 0
_ 6 .0 NE 128IN .0 0 0 2 0 2 !
- B.B
- 7 .0
1—!—!—1—j—I—I—!—!—|—I—I—!—!—[—I—I—I—I—j—I—I—I—I—|—I—I—I—1—j I—I—I—I—j—I—I—t—I—I—I—1—1—I—J—I—I—I—I—{“120 110 100 90 80 70 60 SO 40 30
PPKT I i t J I T
20
App. B, N M R #9: 'H N M R o f ace ty la ted procyanid in B3
(A rom a t ic Region)
SF 400 ,13 SY 230.0 01 0 1 3 4 .ISSI 32766 TD 32766 SW 5613.95 HZ/PT .35
PJACET63.001DATE 17-10-8
PKRDAQneNSTE
5 .01.502 . 6 !
400116297
FW 7300 02 0 . 0 DP 63L PO
LB6BCXCYF lF2HZ/CH PPM/CH Sfl 42
.30 0 . 0
35 .00 0,0 7. 19 4 3 5
32 .53 .06
PJACET63.001DATE 1 7 - ICI-S
SF 400, 13SY 230 .001 6134. ISBI 32766TD 32766SW 5813 95HZ/PT 35
PW 6, 0Aß 1 50AQ 2. 61RS 400NS 116TE 297
FW 730002 0. 00P 63L PO
LB 30GB 0. 0CX 35. 00CY 0. 0F l 3. 82F2 67HZ/CH 36. 01PPM/C* 0»SA 4395. 00
ACETYLATED B3 IN CDCL3
tp. B, NMR #9: 'H N M R of acetylated procyanidin B3
(Aliphat ic Region)
ACETYLATED B3
i—i—i—¡—i—i—r-T—i—i—T—r—i—j—i—i—i—i—i—i—i—i—i—i—i—i—i—i—(—i—r-i—(—I—I—i—i—I 1 r
BBSr9 E
IN CDCL3
OCISOS.150 DATE 17-10-92
SF 400.134 SY 230.0 01 6134.154SI 3276B TD 3276B SN 5B13.953 HZ/PT .355
PKHDAQRGNSTE
6 . 0 1.500 2.BIB
400 116 297
FW 7300 02 0 .0 DP 63L PO
LBGBCXCYFIF2HZ/CMPPM/CM
.3000.0
35.000.07.B32P .511P
83.695 .209
Sfl 4395.00
4 K i n
ACETYLATED B3
kJ IL ^ U II
4 .0
4.5
9 .0
5.5
OC171150.SMX F l PROJ:
PROJH1.001 F2 PROJ:
PROJH1.001 AU PR06:
Z27 ■ AUDATE 1 8 -1 0 -8 2
312 1024S II 512SM2 2B41.17BSKI 1470.581NOO 1
NDW2 SNOMI sSSB2 0SSB1 0MC2 MPLIH RON:FI 7.5S5PF2 3 .746PANO 1COLUMN:FI 7 .623PF2 3 .7 3 2 P
□ 1 1 .9530000PIRSA
9 .0 0
RD 0 .0PN 0 .0DE 2 1 5 .0 0NS 64DS 2DO .0000030P3 4 .5 0NE 128IN .0003400
6.0
6.5
7.0
7.5
App
. B,
NM
R #1
0:
2D CO
SY
of ac
etyl
ated
pr
ocya
nidi
n B
3.
App. B, N M R #11: 'H N M R o f ace ty la ted p rocyan id in C2.
(A ro m a t i c Reg ion)
3F 400 3Y 230.001 b o mSZ 32788 TD 32761 W 1B1B1 H2/PT
p m ano i AQ 1R8 100 MS 187 TC 297
FK 1900002 0 DP 83L P
LB98 0CX 95 CY 0FI 7F2 4HZ/CM 32 PPK/CM Sft 4395
0C1B08.IBDATE 17-1
ACETYLATED C2
OC1509.154 DATC 17-1(
If 400. 8Y 230.0 01 805681 3278«TD 32788 3* 15151HZ/PT
WnoAQR9M8TE
3.1.t.100187287
F* 1900002 0. OP B X PO
LB8B 0CX 39CY 0FI 3F2HZ/C* 39 PPM/C*Sfl 4995
). B, NMR #11: 'H NMR of acetylated procyanidin C2.
(Aliphatic Region)
ACETYLATED C2
PPM
0C150S.150DATE 17-10-92
SF 400.134SY 230.001 6056.473SI 32768TD 32768SH 15151.515HZ/PT .925
PW 3.0RD 1 .000AQ 1.081RG 100NS 4 67TE ^ 9 7
FW 1900002 0.0DP 63L P0
LB .300GB 0.0CX 35.00CY 0.0FI 7.832PF2 . 512PHZ/CM 83.679PPM/CM .209SH 4395.00
1 ! I ] : I I ! \—I—I—I—I—I—i—I—I—I—I—I—I I I—I—i—I—i—r3.0 2.5 2.0 1.5 1.0
OS
fifi
09
S‘9
OO)u»
cuOI
OIo OI
HZDDazo-oíinjaz n ' w o w w n u t ü j M . M .
o o ro <o m o rooi\)í»o oiooii)0> or\)jk< •
cu o o oO b i o
ooo
(0 * (OCD « o o o o o
o
-n*n> n i u r i/juj* xfO»*2M>*rO(/)WOO o ufomcnaracX
oO Xr - oc *ül-ilUN*' IOOWW. . z* ••vil*»’ ■ (Büi Q)li} t* O)
«-Ü)HTfOTJ-D
z wtototnU í V H Ho ^ pü ro
ru m» i* m c n omOIU)A)íw
0 >> “n “noi»NCowon —tf\> n n m m*sjTj»*TjKioa■ x <r 33 oí>*^ V O O D O O ^ acmwcwc*.1 • • • • ■ » --------------
(DO)roaio?U3T\>
A(D
App B, NMR #12: 2D COSY o f acetylated procyanidin C2.
15
jACETYLATEO C2
IN C0CL3
16
B3 IN
ACETONE
App. B, N M R #14: C o m p a r i s o n o f p rocyan id in B3 & isola ted p roduc t .
CAT IN O t t / A C f T O M H i
App. B, NMR # 15(A): In process react ion profile for enzyme reaction.
(+acetone)F i wDATE 14
PJCA.Oi DATI 14
WF AST 2 9 0 . 01 73SI 927 TD 10S n 4ê KZ/PT
NOAOMMSTE Z
m ok 02OP ML
Linex ÏCYFIF2HZ/CM : PPM/CM so ose
App. B, N M R # 1 5(B): Enzyme react ion (+acetone)
■8sæ ær no oO] 73#ii »J>2 lt3* M 4t«;•O/FTH 1S i8 & TC H7S *‘1» ML *
yM Ut».
MA.OATC¡ ■ J pTD ICSM w/rr m
1.0
1:&«IWamTlH "H OP UL ft
I J i i rÇT f.OFt 7.57«F* 4 . MHZ/Ç* » 0 1 !FWCH .0740 9363.00
PJCA.004DATE 19-0
tf 400 SV 290.001 7101 12 S 27M TD 10904 n 4*07 KZ/PT
PH 1m i Ai 1m 20sre oioo02 0. OP M L «
LiM 0.CX *3, Ct 0. PI 7.H . 4. H Z / W 20.fW onIvi 0163.
a s i ö v
If H I127001ÜÍ4
M 4007.1 MZ/PT
l.( 1.1 t.;
20 12 2f7
no
re oioo 0« o,<d p u p o
M 0.(CX 10.( CY 0.1 Pi 7.6 PI, 4.1 HZ/CM 29.1 m u c * .C
“■ 17.4
i '-----1-----'-----1-----'-----1-----*-----r — ' “T- i ' r iB.i
App. B, N M R #15(C ) : Enzym e reac t ion (+ a c e to n e ) T = 6 days
UAf efUfVK
App. B, N M R #16(A): E nzym e reac t ion (N o ace to n e )
PJCA. ooe OATi 1 W - I
» 400 .1•Y 210 ,0 01 eo o * .«a l » 7 8 0 TD 52768 BU 18181.8HZ/FT .1
noA4MmTt
m02OP
u•aaCYF
1.0t.o1.0
20«297
19000 _ 0.0 M L PC
1 .0« . 1 38.0* 0.0
7 . 9 . 4.«
„ Jt.41at .o; .Oi
Ü FFU S.O NO l.W At 1.01 m I«» 78TE 287
FV 10000 02 0 .0 V OLPOLB 1.00
5 98ÌOO6 o .LFI 7.87 F2 4.00 KZ/CM 28.4« FTH/CN .07 8N m i . 00
App. B, N M R #16(B ) : E nzym e reac t ion (No ace tone) .
23
PP
M
PLOTTER TEST
U)T3i‘n'nnn0r": D T > N r u » - » - < x a J c D
nn a. zinCD ID Lüuj \j I m o u io oo i\ jiv ru f\ j0 0 o o o tk LO iv cn a
uj en lu vj
a o "n TDI\> «
cnlu ai r~ ru
oT3 O Oa •
o
—I ZD > IDTJ mwnoot25r\3 »-*
to UJ OîslüJOt-*^to o oCD Oru o
iWHC/iomu] n i d h w ^ -n■D ro—I m*m> ai
ai en en o»-» QJ U) o ■ Au) cd ro cnoo rü o o
«ji r\j O Ll> jùk »-*• •vj
ro
ENZYME REACTION